{"task_id": 1, "task_desc": "Can you assess the damage to buildings after a disaster using satellite images? I have both pre-disaster and post-disaster high-resolution images. Additionally, I would like to understand the types of geospatial objects present in the affected area and get a detailed description of the scene. Please provide a comprehensive analysis of the situation, including any significant changes or damages observed. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif', the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif', the high spatial resolution image for geospatial object segmentation is located at '/data/satellite_images/hsr_image.tif', and the user query is located at '/data/user_queries/user_query.txt'.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/hsr_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 2, "task_desc": "Predict the future water depths during a flood event by analyzing the evolution of precipitation over time. Use the initial radar echo observations to forecast precipitation patterns and convert these predictions into a time-series format suitable for flood depth analysis. The data paths include: 'data/radar_echo_frames/observed_precipitation_data.npy' for radar echo frames, 'data/converted_precipitation/time_series_precipitation.npy' for the converted time-series precipitation data, and 'data/topography/region_dem.npy' for the digital elevation model (DEM) data.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "data/radar_echo_frames/observed_precipitation_data.npy"}, "outputs": ["predicted_radar_echo_frames_path"]}, {"agent": "precipitation_data_convert_tool", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_radar_echo_frames_path"]}, "inputs": {"radar_echo_frames_path": "-0"}, "outputs": ["time_series_array_path"]}, {"agent": "Flood_depth_prediction", "step": 2, "dependence": [1], "dependence_content": {"1": ["time_series_array_path"]}, "inputs": {"region_topography_path": "data/topography/region_dem.npy", "precipitation_data_path": "-1-"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 3, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. The storm has caused significant weather-related degradation in the satellite images we have. Please process these images to restore their clarity and detail, and then identify any visible objects or structures that may have been affected. Additionally, estimate the number of people in crowded areas to prioritize rescue and relief efforts. The satellite image is stored at the following path: /local/data/weather_degraded_images/image1.jpg. These images are satellite captures of the affected urban area, showing various levels of degradation due to the storm.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/weather_degraded_images/image1.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 4, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on infrastructure and natural landscapes. Utilize high-resolution satellite imagery to detect and identify objects such as damaged buildings, fallen trees, and blocked roads in foggy conditions. Provide a detailed analysis of the detected objects, including their locations and the extent of damage, to aid in disaster response and recovery efforts. Additionally, generate a comprehensive report that includes visual descriptions and contextual information about the affected areas to support decision-making for emergency services. The input satellite imagery is stored at 'data/satellite_images/foggy_conditions_image_001.tif'. The user query describing the task is located at 'data/user_queries/disaster_assessment_query.txt'.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/foggy_conditions_image_001.tif"}, "outputs": ["detected_objects_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "data/user_queries/disaster_assessment_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 5, "task_desc": "In the aftermath of a severe weather event, we need to assess the impact on a densely populated urban area. Start by generating high-resolution satellite images of the affected region based on this task description. Once the images are generated, enhance their clarity by restoring any weather-related degradations. Finally, analyze the restored images to accurately count the number of people in the area, which will help in planning and deploying emergency response teams effectively. Please ensure the data path for the initial image generation is correctly set. The data paths are as follows: 'data/captions/urban_area_caption.txt' for the descriptive caption, 'data/metadata/urban_area_metadata.json' for the metadata.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/captions/urban_area_caption.txt", "metadata_path": "data/metadata/urban_area_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 6, "task_desc": "In the aftermath of a severe storm, assess the impact on urban infrastructure by analyzing satellite images. Enhance the clarity of images affected by adverse weather conditions to identify and count the number of people in crowded areas, and detect objects such as vehicles and debris in low-light conditions. Use the path '/data/storm_impact_images/weather_degraded_image_01.jpg' for accessing the necessary images. These images are satellite captures of urban areas affected by the storm, showing various levels of weather degradation and low-light conditions.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/storm_impact_images/weather_degraded_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 7, "task_desc": "In the aftermath of a natural disaster, assess the impact on urban areas by generating high-resolution satellite images using the following description: in the nighttime there are a lot of gathered together as they are affected by the snow storm. Use these images to detect objects in low-light conditions, such as during nighttime or in areas with power outages, to identify critical infrastructure and resources. Additionally, estimate the crowd density in affected regions under adverse weather conditions to aid in efficient resource allocation and emergency response planning. Ensure to provide the path to the necessary data for processing. The necessary data paths are as follows: 'data/captions/disaster_description.txt' for the descriptive caption, 'data/metadata/disaster_metadata.json' for the metadata.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/captions/disaster_description.txt", "metadata_path": "data/metadata/disaster_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 8, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a city by analyzing images captured during the event. The images are affected by heavy rain and low visibility, making it challenging to identify objects and estimate crowd sizes in affected areas. Please process these weather-degraded images to enhance their clarity and detail, then detect any objects present and estimate the number of people in crowded locations. The images are located at the path: /local/data/storm_images/image_001.jpg. These images are captured during the storm and are affected by weather conditions such as heavy rain and low visibility.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/storm_images/image_001.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 9, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on urban infrastructure and public safety. The image is degraded by the weather, and I want to repair it to the high-resolution image. Utilize high-resolution satellite images to detect and count the number of people in crowded areas affected by adverse weather conditions like rain and haze. Additionally, identify any objects or obstacles that may pose a risk to emergency response teams operating in foggy scenarios. This information will aid in coordinating disaster relief efforts and ensuring the safety of affected populations. Please provide the path to the high-resolution satellite images for processing. The paths to the necessary data files are as follows: - Weather degraded image: '/data/images/weather_degraded_image.jpg'", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/images/weather_degraded_image.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 10, "task_desc": "Assess the impact of a recent disaster on an urban area by analyzing high-resolution satellite images taken before and after the event. Identify and categorize any anomalies or damages to buildings and infrastructure, and then classify the type of urban features present in the images. Please provide the path to the pre-disaster and post-disaster high-resolution images for processing. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif' and the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif'.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/pre_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "RGB_GeoImage_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["predicted_category_path"]}], "source": "benchmark"} {"task_id": 11, "task_desc": "Analyze high-resolution satellite images to identify and segment various geospatial objects such as ships, airplanes, and vehicles, and detect anomalies in urban and forest environments. This analysis will help in assessing the impact of potential disaster events, such as wildfires in forests or unexpected changes in urban areas. Please provide the path to the satellite images for processing. The satellite images are stored in the following paths: '/data/satellite_images/geospatial_area_of_interest/image1.tif', '/data/satellite_images/urban_area/image2.tif', '/data/satellite_images/forest_area/image3.tif'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_of_interest/image1.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area/image2.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area/image3.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 12, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a coastal city. The storm has caused significant weather-related degradation in satellite images, making it difficult to identify key infrastructure and areas affected by flooding. Please analyze the high-resolution satellite images to restore clarity and detect any visible objects or structures. Additionally, provide a detailed description of the detected objects and their surroundings to aid in disaster response and recovery efforts. The images are located at the following path: /local/data/weather_degraded_images/image_001.tif. The user query for the task is located at: /local/data/user_queries/query_001.txt.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/weather_degraded_images/image_001.tif"}, "outputs": ["restored_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/local/data/user_queries/query_001.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 13, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or wildfire, assess the extent of damage in a forested area by analyzing high-resolution satellite images. Identify and highlight anomalies such as damaged trees, fallen structures, and new water bodies that may have formed due to the disaster. Provide a detailed contextual description of the affected regions to aid in disaster management and recovery efforts. Please ensure to include the path to the satellite images for analysis. The satellite images can be found at '/data/satellite_images/forest_area_post_disaster.tif'. The user query describing the task is located at '/data/user_queries/disaster_assessment_query.txt'.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_post_disaster.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/disaster_assessment_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 14, "task_desc": "In the aftermath of a severe storm, assess the impact on urban infrastructure by analyzing satellite images. Use the available high-resolution images to detect and identify objects such as damaged buildings, vehicles, and debris in nighttime conditions. Additionally, restore images affected by adverse weather conditions like fog or haze to ensure accurate detection of objects. This analysis will aid in prioritizing emergency response efforts and resource allocation for disaster management. Please provide the path to the satellite images for processing. The satellite image for processing is located at '/data/satellite_images/storm_aftermath/weather_degraded_image_01.jpg' for weather-degraded conditions.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/storm_aftermath/weather_degraded_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 15, "task_desc": "Generate a high-resolution satellite image to assess the impact of a recent natural disaster. Use available metadata and textual descriptions to create an initial low-resolution multi-spectral image of the affected area. Then, enhance this image to a high-resolution RGB format to better analyze the extent of the damage and aid in disaster response planning. Ensure all necessary data inputs are correctly sourced from local directories. The descriptive caption is stored at 'data/input/caption.txt', the metadata is available at 'data/input/metadata.json'.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/input/caption.txt", "metadata_path": "data/input/metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "High-Resolution_Image_Reconstructor", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-", "metadata_path": "data/input/metadata.json"}, "outputs": ["high_resolution_image_path"]}], "source": "benchmark"} {"task_id": 16, "task_desc": "Analyze high-resolution satellite images of urban areas to detect anomalies that may indicate disaster events. Use the path '/local/data/satellite_images/image_urban_area_01.tif' for the input images. The goal is to identify and highlight regions within the urban landscape that deviate from normal patterns, which could suggest the presence of unexpected or irregular features related to disasters. Additionally, classify the detected anomalies to understand the nature of the disaster events and provide insights for effective disaster management and response. The input data includes high-resolution satellite images in TIFF format, which capture detailed urban landscapes, and a binary mask array to specify visible regions during analysis.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/image_urban_area_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "mask_path": "/local/data/satellite_images/mask_urban_area_01.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 17, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. The storm has caused significant weather-related degradation in the satellite images we have. Please restore these images to enhance their clarity and detail. Once the images are restored, estimate the number of people in the affected areas to aid in disaster response and resource allocation. The path to the satellite images is '/local/data/weather_degraded_images/satellite_image_01.jpg'. The restored images will be saved at '/local/data/restored_images/restored_image_01.jpg'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/weather_degraded_images/satellite_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 18, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Identify and classify the extent of building damage, and detect any anomalies in the urban landscape that may indicate areas of concern. Use the following data paths for the analysis: '/local/data/pre_disaster_image.tif' and '/local/data/post_disaster_image.tif'. The pre-disaster image file '/local/data/pre_disaster_image.tif' contains satellite imagery of the area before the disaster and the post-disaster image file '/local/data/post_disaster_image.tif' contains imagery after the event.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/local/data/pre_disaster_image.tif", "post_disaster_image_path": "/local/data/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 19, "task_desc": "Analyze the high-resolution satellite image of the affected region to identify and describe objects and structures that may have been impacted by the recent foggy weather conditions. Provide a detailed contextual description of the scene, highlighting any potential hazards or areas that require immediate attention for disaster management efforts. Please ensure the image is located at the specified path: '/local/data/high_resolution_image.jpg'. The image is a high-resolution satellite capture of the region affected by foggy weather, which will be used to assess the impact on objects and structures. Additionally, the user query describing the task is located at '/local/data/user_query.txt'.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/high_resolution_image.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 20, "task_desc": "Assess the impact of a recent natural disaster on urban infrastructure by analyzing pre- and post-disaster satellite images. Use the images to classify the extent of building damage and generate a detailed damage classification map. Provide a comprehensive contextual description of the affected areas, including potential hazards and recovery needs, to aid in disaster response and management efforts. Ensure all necessary data is correctly referenced in the analysis process. The pre-disaster satellite image is located at '/data/satellite_images/pre_disaster_image.tif', and the post-disaster satellite image is located at '/data/satellite_images/post_disaster_image.tif'. The user query describing the task is located at '/data/user_queries/disaster_assessment_query.txt'.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/disaster_assessment_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 21, "task_desc": "Analyze the satellite image located at '/local/data/satellite_images/satellite_image.jpg' to identify and categorize visible features or objects. Then, provide a detailed contextual description of the identified categories, focusing on their relevance to potential disaster management scenarios, such as infrastructure vulnerability or land use patterns that may impact emergency response efforts. The user query describing the task is located at '/local/data/user_queries/disaster_management_query.txt'.", "structured_plan": [{"agent": "RGB_GeoImage_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/satellite_image.jpg"}, "outputs": ["predicted_category_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_category_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/disaster_management_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 22, "task_desc": "In the aftermath of a natural disaster, it's crucial to assess the impact on both infrastructure and human populations. Using high-resolution satellite images, identify and count the number of people in affected areas, even under challenging weather conditions. Additionally, detect any objects or obstacles that may hinder rescue operations, especially in low-light scenarios. This information will aid in prioritizing rescue efforts and resource allocation. Please provide the path to the satellite images for analysis. The satellite images for analysis are stored at the following paths: '/data/satellite_images/crowd_scene_image_01.jpg' for crowd counting and '/data/satellite_images/low_light_image_01.jpg' for low-light object detection.", "structured_plan": [{"agent": "Crowd_Counting_in_Adverse_Weather", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/crowd_scene_image_01.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 23, "task_desc": "Predict the potential flood depths in a region by analyzing the future precipitation patterns. Use the predicted radar echo frames to convert them into a time-series format suitable for flood depth prediction. Ensure the precipitation data is log-transformed and normalized. Provide the path to the predicted radar echo frames data for processing. The predicted radar echo frames data is located at '/data/predicted_radar_echo_frames/predicted_frames_2023.h5'. The region's topography data is located at '/data/region_topography/dem_array_2023.tif'. The precipitation data is located at '/data/precipitation_data/precipitation_series_2023.csv'.", "structured_plan": [{"agent": "precipitation_data_convert_tool", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/data/predicted_radar_echo_frames/predicted_frames_2023.h5"}, "outputs": ["time_series_array_path"]}, {"agent": "Flood_depth_prediction", "step": 1, "dependence": [0], "dependence_content": {"0": ["time_series_array_path"]}, "inputs": {"region_topography_path": "/data/region_topography/dem_array_2023.tif", "precipitation_data_path": "-0-"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 24, "task_desc": "Assess the impact of a recent disaster on a forested area by analyzing high-resolution satellite images taken before and after the event. Identify any anomalies such as forest fires or other disruptions, and classify the extent of damage to the forest environment. Additionally, detect any changes in the landscape, such as new structures or alterations in the terrain, to support effective disaster management and recovery efforts. Please provide the path to the high-resolution images for analysis. The images are stored locally at the following paths: '/data/satellite_images/pre_disaster_image.tif' for the image taken before the disaster, and '/data/satellite_images/post_disaster_image.tif' for the image taken after the disaster.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/pre_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/data/satellite_images/pre_disaster_image.tif", "image_2_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}], "source": "benchmark"} {"task_id": 25, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. The storm has caused significant weather-related visual degradations, making it challenging to accurately detect objects and count the number of people in the affected zones. Please process the available images to enhance their quality, enabling reliable object detection and crowd counting. This will help in coordinating emergency response efforts and ensuring the safety of the residents. Provide the path to the images that need to be analyzed. The images are stored locally and can be accessed at the following paths: '/data/storm_impact/low_light_images/image1.jpg' for low-light object detection and '/data/storm_impact/crowd_images/image2.jpg' for crowd counting in adverse weather conditions.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/storm_impact/low_light_images/image1.jpg"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/storm_impact/crowd_images/image2.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 26, "task_desc": "Can you analyze the remote sensing images to identify and highlight areas that are prone to landslides? Once identified, provide a detailed interpretation of the scene, including any anomalies or unusual patterns that could indicate potential risks or require further investigation. Please ensure the analysis is comprehensive and considers all relevant geospatial features. The high-resolution aerial or satellite image capturing mountainous terrain is located at '/data/remote_sensing/landslide_analysis/mountainous_terrain_image.tif'. The user query describing the task is located at '/data/remote_sensing/landslide_analysis/user_query.txt'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/remote_sensing/landslide_analysis/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/remote_sensing/landslide_analysis/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 27, "task_desc": "Analyze the satellite imagery to identify and classify areas affected by a recent natural disaster, such as a flood or wildfire. Provide a detailed description of the scene, including any visible damage to infrastructure or changes in the landscape. Use the high-resolution images available to ensure accurate and comprehensive scene interpretation. Please include the path to the satellite image data in your analysis. The satellite image data is located at '/data/satellite_images/sentinel2_image.tif'. The binary mask array is located at '/data/masks/sentinel2_mask.npy'. The user query file is located at '/data/queries/user_query.txt'.", "structured_plan": [{"agent": "Multi_Spectral_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_image.tif", "mask_path": "/data/masks/sentinel2_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["reconstructed_image_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/queries/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 28, "task_desc": "In the aftermath of a recent disaster, we need to assess the impact on urban areas by generating high-resolution images of the affected locations. Using available low-resolution images from before and after the disaster, along with a high-resolution image from before the disaster, create a high-resolution image of the area post-disaster. Once the high-resolution post-disaster image is generated, analyze it to detect any anomalies or irregularities in the urban landscape that may indicate damage or changes caused by the disaster. This will help in identifying areas that require immediate attention and aid in planning recovery efforts. Please provide the paths to the low-resolution images from before and after the disaster, as well as the high-resolution image from before the disaster. The paths are as follows: 'data/low_res_image_before_disaster.jpg', 'data/low_res_image_after_disaster.jpg', 'data/high_res_image_before_disaster.jpg', and 'data/metadata.json'. The low-resolution images are satellite images taken before and after the disaster, while the high-resolution image is a satellite image taken before the disaster. The metadata file contains information such as latitude, longitude, ground sampling distance (GSD), cloud cover fraction, year, month, and day.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/low_res_image_before_disaster.jpg", "high_res_image_T_path": "data/high_res_image_before_disaster.jpg", "low_res_image_T_prime_path": "data/low_res_image_after_disaster.jpg", "metadata_path": "data/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 29, "task_desc": "In the aftermath of a severe weather event, such as a foggy day, we need to assess the impact on urban areas. Utilize high-resolution satellite images to detect objects and count crowds in affected regions. This will help in understanding the extent of the disaster, identifying areas with high population density, and planning effective relief operations. Please provide the path to the satellite images for analysis. The satellite images are stored in the local machine at the following paths: '/data/satellite_images/foggy_day_image_01.jpg' for foggy weather conditions and '/data/satellite_images/normal_day_image_01.jpg' for normal weather conditions.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/foggy_day_image_01.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/normal_day_image_01.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 30, "task_desc": "Analyze the provided high-resolution satellite imagery to identify and classify geospatial objects such as buildings, vehicles, and natural features. Generate a detailed contextual description of the scene, highlighting any anomalies or potential disaster-related changes, such as landslides or urban disturbances. Ensure the analysis includes a comprehensive interpretation of the scene to aid in disaster management efforts. The high-resolution satellite imagery is located at '/data/satellite_images/high_res_image_01.tif'. The user query describing the task is located at '/data/user_queries/query_01.txt'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/high_res_image_01.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/query_01.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 31, "task_desc": "Analyze the potential impact of a recent wildfire in a forested area by using satellite imagery. Start by enhancing the resolution of the available low-resolution multi-spectral satellite images to obtain high-resolution images. Then, identify and map any anomalies in the forest environment that may indicate the presence of wildfire damage or other disruptions. Provide a detailed anomaly map highlighting areas of concern for further investigation and response planning. Please ensure the path to the low-resolution multi-spectral image is correctly specified. The low-resolution multi-spectral satellite image is located at '/data/satellite_images/low_res_image.tif'. The metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image.tif", "metadata_path": "/data/satellite_images/metadata.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 32, "task_desc": "In the aftermath of a severe storm, we need to assess the damage to infrastructure and identify any potential hazards. The storm has left the area covered in fog and other weather-related obstructions, making it difficult to see clearly. Please process the images captured during the storm to remove these weather-related degradations and enhance their clarity. Once the images are restored, perform object detection to identify any damaged structures, fallen trees, or other hazards that need immediate attention. The image is located at the path: '/local/data/storm_images/storm_image_001.jpg'. These image is taken during the storm and contain various weather-related obstructions such as fog and rain.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/storm_images/storm_image_001.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 33, "task_desc": "Identify and analyze potential landslide-prone areas using satellite imagery. Utilize the available low-resolution multi-spectral images to generate high-resolution images, and then detect and highlight regions that show signs of landslide anomalies. This will help in assessing the risk and planning for disaster management in mountainous regions. Please provide the path to the low-resolution multi-spectral image data. The low-resolution multi-spectral image data can be found at '/data/satellite_images/low_res/multi_spectral_image_01.tif'. The metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata/contextual_details_01.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res/multi_spectral_image_01.tif", "metadata_path": "/data/satellite_images/metadata/contextual_details_01.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 34, "task_desc": "In the aftermath of a natural disaster, we need to assess the extent of damage and changes in the affected region using satellite imagery. Please analyze the temporal sequence of satellite images to classify the type of disaster and generate a detailed report on the damage levels, changes over time, and the types of geospatial objects affected. This information will be crucial for coordinating relief efforts and planning reconstruction. Ensure to provide a comprehensive interpretation of the scene, highlighting significant modifications and damage classifications. The temporal sequence of satellite images is stored in the local machine at the following path: '/data/satellite_images/image_sequence_01.png'. The user query describing the task is located at '/data/user_queries/disaster_assessment_query.txt'.", "structured_plan": [{"agent": "Temporal_Image_Sequence_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/image_sequence_01.png"}, "outputs": ["classification_results_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["classification_results_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/disaster_assessment_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 35, "task_desc": "Can you analyze the satellite images of the urban area to identify any anomalies or irregularities that might indicate potential damage or unusual patterns? Please provide a detailed description of the findings, including the severity of any detected damage, using the images located at '/local/data/satellite_images/urban_area_image_01.jpg'. The user query is located at '/local/data/user_queries/query_01.txt'.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/urban_area_image_01.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/query_01.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 36, "task_desc": "In the event of a flood, assess the impact on traffic flow by predicting water depths across a region and determining how these depths affect traffic speeds on road networks. Use the available topography and precipitation data to simulate the flood scenario and analyze the resulting traffic conditions. Ensure that the precipitation data is properly formatted and normalized before use. Provide insights on potential traffic disruptions and suggest alternative routes if necessary. The data paths are as follows: 'data/topography/region_dem.tif' for the digital elevation model, 'data/precipitation/precipitation_data.csv' for the precipitation data, 'data/road_network/road_network.shp' for the road network, and 'data/flood_depth/flood_depth_data.csv' for the flood inundation depth.", "structured_plan": [{"agent": "Flood_depth_prediction", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"region_topography_path": "data/topography/region_dem.tif", "precipitation_data_path": "data/precipitation/precipitation_data.csv"}, "outputs": ["predicted_water_depths_path"]}, {"agent": "depth_speed_model", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_water_depths_path"]}, "inputs": {"road_network_path": "data/road_network/road_network.shp", "flood_depth_path": "-0-"}, "outputs": ["traffic_speed_path"]}], "source": "benchmark"} {"task_id": 37, "task_desc": "In the aftermath of a forest disaster, such as a forest fire, analyze high-resolution satellite images taken before and after the event to assess the impact. Utilize the images to detect and map anomalies in the forest environment, identifying areas with significant changes or deviations from the normal forest pattern. Additionally, perform geospatial object segmentation to identify and classify objects within the affected area, such as vehicles or structures, which may have been impacted by the disaster. This comprehensive analysis will aid in understanding the extent of the damage and support recovery and management efforts. Please provide the path to the high-resolution images for processing. The paths to the required data are as follows: 'data/satellite_images/before_event_image.tif' for the image taken before the event, 'data/satellite_images/after_event_image.tif' for the image taken after the event", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/before_event_image.tif", "image_2_path": "data/satellite_images/after_event_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 38, "task_desc": "Analyze the provided high-resolution satellite images to identify and describe any anomalies or unusual patterns in forest and urban environments that could indicate potential disaster risks. Provide a comprehensive interpretation of the scene, highlighting areas of concern and potential threats to aid in disaster management and response planning. The data paths for the images and user queries are as follows: - Forest area image: /local/data/satellite_images/forest_area_image.tif - Urban area image: /local/data/satellite_images/urban_area_image.tif - User query for forest analysis: /local/data/user_queries/forest_query.txt - User query for urban analysis: /local/data/user_queries/urban_query.txt", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/forest_query.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/local/data/user_queries/urban_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 39, "task_desc": "In the aftermath of a severe weather event, assess the impact on urban infrastructure by analyzing satellite imagery. Begin by generating high-resolution satellite images based on available metadata and textual prompts. Enhance these images to remove any weather-related degradations such as rain, snow, or haze. Once the images are restored, perform object detection to identify and locate critical infrastructure elements that may have been affected. Additionally, estimate the crowd density in public areas to understand the movement and gathering of people in response to the disaster. This comprehensive analysis will aid in effective disaster management and resource allocation. The data paths for this task are as follows: 'data/captions/context_caption.txt' for the descriptive caption, 'data/metadata/metadata_vector.json' for the metadata, 'data/images/weather_degraded_image.jpg' for the weather-degraded image, 'data/images/restored_image.jpg' for the restored image, and 'data/images/crowd_scene.jpg' for the crowd scene image.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/captions/context_caption.txt", "metadata_path": "data/metadata/metadata_vector.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 3, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 40, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on urban infrastructure and public safety. Generate satellite imagery to detect and count the number of people in affected areas, even under challenging weather conditions like fog or low light. Additionally, identify any visible objects such as vehicles or debris that may pose hazards. Please use the '/data/captions/storm_aftermath_caption.txt' for descriptive captions and '/data/metadata/storm_aftermath_metadata.json' for metadata information.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/data/captions/storm_aftermath_caption.txt", "metadata_path": "/data/metadata/storm_aftermath_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 3, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 41, "task_desc": "Can you help me identify and highlight the areas that have undergone significant changes between two sets of satellite images taken before and after a natural disaster? I have the images stored at '/local/data/pre_disaster_images/image1.tif' and '/local/data/post_disaster_images/image2.tif'. I need a detailed change map to assess the damage accurately.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/local/data/pre_disaster_images/image1.tif", "image_2_path": "/local/data/post_disaster_images/image2.tif"}, "outputs": ["change_map_path"]}], "source": "benchmark"} {"task_id": 42, "task_desc": "Can you help identify and count the number of emergency vehicles present in images taken during a nighttime disaster response operation? The specific image file is '/data/disaster_night_images/image1.jpg'. These images are captured under low-light conditions and require specialized detection models to identify emergency vehicles.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/disaster_night_images/image1.jpg"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 43, "task_desc": "I have some low-resolution satellite images of areas affected by a recent natural disaster. Can you help me generate high-resolution RGB images from these to better assess the damage? The images are located at '/local/data/satellite_images/low_res_image_01.tif'. The metadata providing contextual details for the satellite images is located at '/local/data/satellite_images/metadata_01.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/low_res_image_01.tif", "metadata_path": "/local/data/satellite_images/metadata_01.json"}, "outputs": ["high_resolution_image_path"]}], "source": "benchmark"} {"task_id": 44, "task_desc": "Can you generate a satellite image depicting the current state of a wildfire-affected region, using metadata and any available textual information to illustrate the extent of the damage? The descriptive caption that provides context for the image to be generated can be found at '/local/data/wildfire_caption.txt'. The normalized numerical metadata vector providing additional contextual information to guide the image generation is located at '/local/data/wildfire_metadata.json'.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/local/data/wildfire_caption.txt", "metadata_path": "/local/data/wildfire_metadata.json"}, "outputs": ["generated_image_path"]}], "source": "benchmark"} {"task_id": 45, "task_desc": "I have a collection of satellite images from a recent natural disaster. Can you help me identify and categorize the different types of damage visible in these images? The specific images to be analyzed is '/data/disaster_images/image1.jpg'", "structured_plan": [{"agent": "RGB_GeoImage_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/disaster_images/image1.jpg"}, "outputs": ["predicted_category_path"]}], "source": "benchmark"} {"task_id": 46, "task_desc": "I have a set of surveillance videos from a recent natural disaster event. Can you analyze these videos to identify any unusual activities or anomalies that occurred during the event? Please provide both a general overview of any anomalies detected and specific categories of these anomalies if possible. The videos are located at /local/data/disaster_videos/video1.mp4, /local/data/disaster_videos/video2.mp4, and /local/data/disaster_videos/video3.mp4. The extracted features for analysis are stored at /local/data/extracted_features/video1_features.csv, /local/data/extracted_features/video2_features.csv, and /local/data/extracted_features/video3_features.csv.", "structured_plan": [{"agent": "Video_anomaly_detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"extracted_video_feature_path": "/local/data/extracted_features/video1_features.csv"}, "outputs": ["coarse_grained_anomaly_confidence_path", "fine_grained_anomaly_map_path"]}, {"agent": "Video_anomaly_detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"extracted_video_feature_path": "/local/data/extracted_features/video2_features.csv"}, "outputs": ["coarse_grained_anomaly_confidence_path", "fine_grained_anomaly_map_path"]}, {"agent": "Video_anomaly_detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"extracted_video_feature_path": "/local/data/extracted_features/video3_features.csv"}, "outputs": ["coarse_grained_anomaly_confidence_path", "fine_grained_anomaly_map_path"]}], "source": "benchmark"} {"task_id": 47, "task_desc": "Can you analyze the population mobility trends during the recovery phase after a disaster using the data from '/data/disaster_recovery/mobility_data.csv'? I need insights on how the population is moving within and between different regions to help with planning and resource allocation. Additionally, use the initial abnormal population mobility data from '/data/disaster_recovery/initial_abnormal_mobility_graph.json' and the normal population mobility data from '/data/disaster_recovery/normal_mobility_graph.json'. These files contain graph objects that represent the state of population mobility immediately after the disaster and prior to the disaster, respectively.", "structured_plan": [{"agent": "Post_Disaster_Mobility_Recovery", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"initial_abnormal_population_mobility_graph_path": "/data/disaster_recovery/initial_abnormal_mobility_graph.json", "normal_population_mobility_graph_path": "/data/disaster_recovery/normal_mobility_graph.json"}, "outputs": ["predicted_population_mobility_graph_path"]}], "source": "benchmark"} {"task_id": 48, "task_desc": "Can you help me identify and extract all the place names mentioned in the disaster-related reports I have? The reports are located at '/data/disaster_reports/report1.txt', '/data/disaster_reports/report2.txt', and '/data/disaster_reports/report3.txt'. These reports contain detailed descriptions of various disaster events and their impacts, which include numerous place names that need to be extracted.", "structured_plan": [{"agent": "Toponym_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"sentence": "/data/disaster_reports/report1.txt"}, "outputs": ["detected_toponyms_path"]}, {"agent": "Toponym_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"sentence": "/data/disaster_reports/report2.txt"}, "outputs": ["detected_toponyms_path"]}, {"agent": "Toponym_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"sentence": "/data/disaster_reports/report3.txt"}, "outputs": ["detected_toponyms_path"]}], "source": "benchmark"} {"task_id": 49, "task_desc": "Can you predict the changes in transportation demand and supply during a natural disaster event using the data from '/data/disaster_mobility_patterns.csv', '/data/historical_mobility_data_tensor.csv', and '/data/temporal_covariates_matrix.csv'? I need to understand how different transport modes will be affected and interact with each other during the event. The file '/data/disaster_mobility_patterns.csv' contains data on mobility patterns during past disaster events. The file '/data/historical_mobility_data_tensor.csv' includes historical multimodal mobility data, and '/data/temporal_covariates_matrix.csv' provides auxiliary temporal information such as time-of-day, day-of-week, whether it is a holiday, and other temporal features.", "structured_plan": [{"agent": "Multimodal_mobility_prediction_under_events", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"mobility_tensor_path": "/data/historical_mobility_data_tensor.csv", "temporal_covariates_path": "/data/temporal_covariates_matrix.csv"}, "outputs": ["predicted_mobility_path"]}], "source": "benchmark"} {"task_id": 50, "task_desc": "Can you identify and categorize different types of natural disaster events from the recent news article I have, even if there are no prior examples of these specific disaster types in the data? The news article is stored in the local machine at the following path: '/data/news_articles/article1.txt'. These files contain text data of recent news articles related to various natural disasters. You should use the prompt located at '/prompts/prompt1.txt' and trigger located at the 'triggers/trigger.npy'.", "structured_plan": [{"agent": "Event_detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"context": "/data/news_articles/article1.txt", "prompt": "/prompts/prompt1.txt", "trigger": "triggers/trigger.npy"}, "outputs": ["event_type_path", "trigger_positions_path", "confidence_scores_path"]}], "source": "benchmark"} {"task_id": 51, "task_desc": "Analyze the provided high-resolution satellite image to identify and segment various geospatial objects, such as vehicles and buildings, and detect any anomalies that may indicate potential disaster events, including landslides or urban disruptions. The data paths for the required images and preprocessing requirements are as follows: 'data/satellite_images/geospatial_area_image.tif' for the geospatial area of interest, 'data/satellite_images/mountainous_terrain_image.tif' for the mountainous terrain, and 'data/satellite_images/urban_area_image.tif' for the urban area.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/geospatial_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 52, "task_desc": "I have an image located at '/data/satellite_images/forest_area_image.tif'. Analyze the provided high-resolution satellite image to identify and classify potential anomalies in a forest environment, such as signs of wildfires or other disruptions. Additionally, I have an image located in located at '/data/satellite_images/mountainous_terrain_image.tif', you need to assess the image for any landslide-prone areas and finally I have the other image in '/data/satellite_images/geospatial_area_image.tif' please segment various geospatial objects like vehicles or structures.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 53, "task_desc": "In the aftermath of a foggy storm, assess the impact on urban infrastructure and population density. Utilize high-resolution satellite imagery to detect objects affected by the weather conditions, and estimate the crowd density in affected areas. Provide a comprehensive report that includes description of the scene, highlighting key objects and their conditions, as well as an estimation of the crowd count in the region. The high-resolution satellite imagery data can be found at '/data/satellite_images/foggy_storm_image_01.tif'. User query is located at '/data/user_queries/queries.txt'", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/foggy_storm_image_01.tif"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/queries.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 54, "task_desc": "I need to analyze satellite imagery to identify potential disaster-related anomalies in various environments. Specifically, I want to detect anomalies in forest areas that could indicate issues like wildfires or other disruptions. Additionally, I need to identify landslide-prone areas in mountainous regions and detect unusual patterns in urban settings that might suggest disaster events. Please use the available satellite images and provide detailed anomaly maps for each scenario. The data is located at the following paths: '/data/satellite_images/forest_area_low_res.tif' for the 1st task, '/data/satellite_images/mountainous_area_high_res.tif' for the 2nd task, '/data/satellite_images/urban_area_high_res.tif' for the last.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_low_res.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_area_high_res.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_high_res.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 55, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or earthquake, it is crucial to assess the impact on infrastructure and population. Using high-resolution satellite images, identify and count the number of people in crowded areas. Detect objects that may indicate damage or obstruction in the dark environment. I also have images in the fog environment which needs to detect objects in the image. This information will aid in prioritizing emergency response efforts and allocating resources effectively. The data paths for the images are as follows: - High-resolution satellite images for count the number of people in crowded areas: '/data/satellite_images/crowd_scene_image_01.jpg'- Images damage or obstruction in the dark environment: '/data/satellite_images/low_light_image_01.jpg'- Imagesthe fog environment which needs to detect objects in the image: '/data/satellite_images/foggy_image_01.jpg'.", "structured_plan": [{"agent": "Crowd_Counting_in_Adverse_Weather", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/crowd_scene_image_01.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/foggy_image_01.jpg"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 56, "task_desc": "Analyze the provided high-resolution satellite image to identify and map potential landslide-prone areas. Additionally, detect any anomalies in urban environments that may indicate disaster events, and segment various geospatial objects within the image for a comprehensive understanding of the scene. Please ensure the image is related to a landslide scenario and urban disaster events. The path to the image data is required for processing. The image data is located at '/data/satellite_images/landslide_scenario_image.tif' for landslide analysis, '/data/satellite_images/urban_disaster_image.tif' for urban anomaly detection, and '/data/satellite_images/geospatial_area_image.tif' for geospatial object segmentation.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/landslide_scenario_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 57, "task_desc": "Analyze the satellite imagery to identify and describe areas affected by potential landslides and urban anomalies. Provide a detailed interpretation of the scene, highlighting any deviations from typical patterns that could indicate disaster-prone regions. Use the data to generate a comprehensive report on the current state of the landscape, focusing on potential risks and anomalies. The data includes high-resolution aerial or satellite images capturing mountainous terrain and urban areas. The paths to these data files are as follows: 'data/satellite_images/mountainous_terrain_image.jpg', 'data/satellite_images/urban_area_image.jpg'. The user queries are stored at 'data/user_queries/user_queries.txt'", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/mountainous_terrain_image.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/urban_area_image.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "data/user_queries/user_queries.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "data/user_queries/user_queries.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 58, "task_desc": "Analyze the potential impact of natural disasters in forested and mountainous regions by processing satellite imagery. Utilize the low-resolution multi-spectral satellite images to generate high-resolution images, and then identify anomalies such as man-made structures, water bodies, and signs of forest health issues in forest areas. Additionally, detect landslide-prone or affected areas in mountainous terrains. The low-resolution multi-spectral image data can be found at '/data/satellite_images/low_res/multi_spectral_image_01.tif'. The metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata/contextual_metadata_01.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res/multi_spectral_image_01.tif", "metadata_path": "/data/satellite_images/metadata/contextual_metadata_01.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 59, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Identify and classify any anomalies or changes in the urban landscape, such as damaged buildings or altered infrastructure. Additionally, evaluate the potential for landslides in the affected area by detecting any deviations from typical terrain patterns. Use the following data paths for the analysis: '/data/satellite_images/pre_disaster_image.tif', '/data/satellite_images/post_disaster_image.tif', '/data/terrain_images/mountainous_area_image.tif'. The pre-disaster and post-disaster images are high-resolution satellite images of the urban area, while the mountainous area image is used for landslide evaluation.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/terrain_images/mountainous_area_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 60, "task_desc": "Analyze a high-resolution satellite image to identify and segment various geospatial objects, such as buildings, vehicles, and natural features, in an urban area affected by a recent disaster. Additionally, detect and highlight any anomalies that deviate from typical urban patterns, which could indicate damage or irregularities caused by the disaster. Provide a detailed contextual description of the scene, including potential risks and areas requiring immediate attention. The image is located at '/local/data/satellite_image/high_res_image.tif'.The user query for the task is located at '/local/data/user_queries/query.txt'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_image/high_res_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/local/data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 61, "task_desc": "Can you analyze the satellite images to assess the damage caused by a recent disaster? I have high-resolution images from before and after the event. Please generate a detailed damage classification map and provide a comprehensive contextual description of the affected areas, including any significant geospatial objects or anomalies detected. The images are stored in the following paths: '/local/data/satellite_images/pre_disaster_image.tif', '/local/data/satellite_images/post_disaster_image.tif'. The user query is located at '/local/data/user_queries/user_query.txt'", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/local/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/local/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/post_disaster_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/local/data/user_queries/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 62, "task_desc": "In the aftermath of a severe storm, we need to assess the damage in a foggy and low-light environment. Utilize the available tools to enhance and restore images captured during the storm, which are degraded by rain and fog. Once the images are restored, detect and identify any objects or obstacles that may pose a risk to emergency response teams. This will help in planning safe and efficient routes for rescue operations. Please ensure the path to the weather-degraded images is correctly set for processing. The weather-degraded images are stored at '/local/data/weather_degraded_images/storm_image_01.jpg'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/weather_degraded_images/storm_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}, {"agent": "Low-Light_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 63, "task_desc": "In the aftermath of a natural disaster, assess the impact on urban and forest environments by analyzing high-resolution satellite images taken before and after the event. Utilize the available data to identify anomalies in urban areas, such as damaged infrastructure or unexpected changes, and detect deviations in forest regions, like signs of forest fires or other disturbances. Additionally, evaluate landslide-prone areas in mountainous regions by examining changes in terrain patterns. Provide detailed anomaly maps for each environment to aid in disaster response and recovery efforts. The data paths for the images are as follows: - Urban area satellite image after the event: /data/satellite_images/urban/after_event_urban_image.tif - Forest area satellite image after the event: /data/satellite_images/forest/after_event_forest_image.tif - Mountainous terrain image for landslide analysis: /data/satellite_images/mountainous/landslide_analysis_image.tif", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban/after_event_urban_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest/after_event_forest_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous/landslide_analysis_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 64, "task_desc": "In the aftermath of a natural disaster, we need to assess the impact on infrastructure and categorize the affected areas. Using high-resolution satellite images taken before and after the disaster, generate a high-resolution image of the affected area post-disaster. Then, classify the generated high-resolution image to identify and categorize the types of damage or changes that have occurred. Additionally, analyze a sequence of temporal satellite images to understand the progression of the disaster's impact over time. Please ensure the data paths for the required images are correctly set up for processing. The data paths are as follows: - Low-resolution satellite image at time T: '/data/satellite_images/low_res_image_T.jpg'- High-resolution satellite image at time T: '/data/satellite_images/high_res_image_T.jpg'- Low-resolution satellite image at time T': '/data/satellite_images/low_res_image_T_prime.jpg'- Metadata associated with the satellite images: '/data/satellite_images/metadata.json'- Temporal sequence of satellite images: '/data/satellite_images/temporal_sequence/temporal_sequence.npy'", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/data/satellite_images/low_res_image_T.jpg", "high_res_image_T_path": "/data/satellite_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "/data/satellite_images/low_res_image_T_prime.jpg", "metadata_path": "/data/satellite_images/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "RGB_GeoImage_Classifier", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["predicted_category_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence/temporal_sequence.npy"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 65, "task_desc": "Predict the future water depths in a region during a flood event by analyzing the initial radar echo frames of precipitation. Use the radar data to forecast precipitation intensity and convert it into a format suitable for flood depth prediction. Ensure the predictions account for spatial and temporal variability in rainfall to aid in disaster management and response planning. Please provide the path to the initial radar echo frames data. The initial radar echo frames data can be found at '/data/radar/initial_radar_echo_frames_2023_10_01.dat'. The rasterized digital elevation model (DEM) array is located at '/data/topography/region_dem_2023_10_01.tif'.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/data/radar/initial_radar_echo_frames_2023_10_01.dat"}, "outputs": ["predicted_radar_echo_frames_path"]}, {"agent": "precipitation_data_convert_tool", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_radar_echo_frames_path"]}, "inputs": {"radar_echo_frames_path": "-0-"}, "outputs": ["time_series_array_path"]}, {"agent": "Flood_depth_prediction", "step": 2, "dependence": [1], "dependence_content": {"1": ["time_series_array_path"]}, "inputs": {"region_topography_path": "/data/topography/region_dem_2023_10_01.tif", "precipitation_data_path": "-1-"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 66, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Generate a change map to identify areas of significant alteration, such as new constructions or damaged buildings. Further classify the extent of building damage into categories like no damage, minor damage, major damage, or destroyed. Additionally, detect any anomalies in the urban landscape that may indicate irregularities or unexpected changes due to the disaster. Ensure to use the appropriate data paths for the pre- and post-disaster images. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif', the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif', and the high-resolution image for anomaly detection is located at '/data/satellite_images/high_res_image.tif'.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/data/satellite_images/pre_disaster_image.tif", "image_2_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/high_res_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 67, "task_desc": "Analyze high-resolution satellite images to identify and map anomalies in both urban and forest environments that could indicate potential disaster events. For urban areas, focus on detecting irregularities that may suggest infrastructure damage or unexpected changes due to disasters. In forest regions, identify anomalies that could signal environmental threats such as wildfires or disease outbreaks. Additionally, classify the detected features to better understand the nature of these anomalies and their potential impact on disaster management efforts. Please provide the path to the satellite image data for processing. The data paths are as follows: - Urban satellite image: '/data/satellite_images/urban_area_high_res.tif'- Forest satellite image: '/data/satellite_images/forest_area_high_res.tif'- Multi-spectral satellite image: '/data/satellite_images/sentinel2_multispectral.tif'- Binary mask for multispectral image: '/data/masks/sentinel2_mask.npy'", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_high_res.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_high_res.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_multispectral.tif", "mask_path": "/data/masks/sentinel2_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 68, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a coastal city. The storm has caused significant weather-related degradation in the images captured by surveillance cameras, making it difficult to identify objects and estimate crowd density in affected areas. Please process the weather-degraded images to restore clarity and detail, then detect any objects present and estimate the crowd count in the restored images. This information will be crucial for coordinating emergency response efforts and ensuring public safety. The path to the weather-degraded images is: /local/data/weather_degraded_images/image_set_01.jpg.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/weather_degraded_images/image_set_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 69, "task_desc": "Generate a high-resolution satellite image of a disaster-affected area using the following descriptions: the infrastructure damaged by car accident in the nighttime and the foggy environments. Once the image is generated, analyze it to detect objects in low-light conditions and identify any obstructions or damages caused by foggy weather conditions. This will help in assessing the impact of the disaster and planning the necessary relief operations. The metadata and textual prompts are stored in the following paths: 'data/metadata/metadata_vector.json' and 'data/text_prompts/descriptive_caption.txt'.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/text_prompts/descriptive_caption.txt", "metadata_path": "data/metadata/metadata_vector.json"}, "outputs": ["generated_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 70, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or earthquake, it is crucial to assess the impact on infrastructure and human presence in affected areas. Utilize available satellite imagery to detect and identify objects in the dark environments and foggy environments separately to understand the extent of damage. Additionally, evaluate the crowd density in these areas to aid in efficient resource allocation and rescue operations. The data paths for the images are as follows: 'data/low_light_images/image1.jpg' for low-light conditions, 'data/foggy_images/image2.jpg' for foggy conditions, and 'data/crowd_images/image3.jpg' for crowd scenes.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/low_light_images/image1.jpg"}, "outputs": ["detections_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/foggy_images/image2.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/crowd_images/image3.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 71, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a coastal city. Utilize high-resolution satellite imagery to detect and identify objects in foggy conditions, such as damaged infrastructure and vehicles. Additionally, provide a detailed contextual description of the detected objects to aid in emergency response planning. Ensure the analysis accounts for low visibility and adverse weather conditions to deliver accurate and actionable insights for disaster management teams. The input data includes high-resolution satellite images stored at 'data/satellite_images/foggy_conditions_image_001.jpg' and user queries stored at 'data/user_queries/emergency_response_query.txt'.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/foggy_conditions_image_001.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "data/user_queries/emergency_response_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 72, "task_desc": "In the aftermath of a natural disaster, assess the impact on a specific location by generating high-resolution images of the area based on images before and after the event. Utilize these images to identify and categorize geospatial objects affected by the disaster, such as buildings, vehicles, and infrastructure. Additionally, detect and map any landslide-prone or affected areas to aid in disaster response and recovery efforts. Ensure to provide the path to the necessary satellite images for processing. The paths to the necessary data are as follows: 'data/satellite_images/low_res_image_T.jpg' for the low-resolution satellite image at time T, 'data/satellite_images/high_res_image_T.jpg' for the high-resolution satellite image at time T, 'data/satellite_images/low_res_image_T_prime.jpg' for the low-resolution satellite image at time T', and 'data/metadata/metadata.json' for the metadata associated with the satellite images.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/satellite_images/low_res_image_T.jpg", "high_res_image_T_path": "data/satellite_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "data/satellite_images/low_res_image_T_prime.jpg", "metadata_path": "data/metadata/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 73, "task_desc": "Generate high-resolution satellite images using the following description: a lot of people gathered together due to snow storm and the image should be weather degraded. to assess the impact of a recent natural disaster. Enhance the generated images to restore clarity and detail affected by adverse weather conditions. Utilize the restored images to accurately count the number of people in affected areas, ensuring reliable crowd estimation for effective disaster response and management. The data paths are as follows: 'caption_path': '/local/data/captions/snow_storm_caption.txt', 'metadata_path': '/local/data/metadata/snow_storm_metadata.json'", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/local/data/captions/snow_storm_caption.txt", "metadata_path": "/local/data/metadata/snow_storm_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 74, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or flood, it is crucial to assess the extent of damage and identify areas in need of immediate assistance. Utilize satellite imagery to detect and analyze objects in affected regions, even under challenging conditions like fog or low light. This will help in identifying critical infrastructure, such as roads and bridges, that may be obstructed or damaged, and ensure efficient allocation of resources for rescue and recovery operations. Please provide the path to the satellite image data for processing. The satellite image data for processing can be found at the following paths: '/data/satellite_images/foggy_conditions/image1.png' for images captured in foggy or normal weather conditions, and '/data/satellite_images/low_light_conditions/image2.png' for images captured under low-light conditions.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/foggy_conditions/image1.png"}, "outputs": ["detected_objects_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_conditions/image2.png"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 75, "task_desc": "Analyze the satellite imagery to identify and describe potential urban anomalies. Provide a detailed interpretation of the scene, highlighting any unusual patterns or features that could indicate a risk of disaster. The data includes high-resolution satellite images and user queries that describe the specific tasks to be performed. The satellite image is stored at 'data/satellite_images/urban_area_high_res.png', the user queries are stored at 'data/queries/user_queries.txt'", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/urban_area_high_res.png"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "data/queries/user_queries.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 76, "task_desc": "Analyze satellite imagery to identify potential landslide-prone areas. Use high-resolution images to detect anomalies in forested regions that may indicate environmental changes or potential disasters. Additionally, evaluate urban areas for unexpected features that could signal disaster events. Ensure to include the path to the satellite imagery data for processing. The satellite imagery data is stored in the following paths: '/data/satellite_images/mountainous_terrain_image.tif' for mountainous terrain, '/data/satellite_images/forest_area_image.tif' for forest areas, and '/data/satellite_images/urban_area_image.tif' for urban areas.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 77, "task_desc": "Analyze the satellite imagery to identify and describe potential landslide-prone areas and anomalies in forest environments. Provide a detailed interpretation of the scene, highlighting any deviations from typical patterns, such as diseased trees, man-made structures, or water bodies. Ensure the analysis includes a comprehensive understanding of the geospatial context and potential risks associated with these anomalies. Please include the path to the satellite imagery data for processing. The satellite imagery data is located at '/data/satellite_images/mountainous_terrain_image.tif' for landslide analysis and '/data/satellite_images/forest_area_image.tif' for forest anomaly detection. The user query describing the task is located at '/data/user_queries/task_description.txt'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/task_description.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/task_description.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 78, "task_desc": "Analyze high-resolution satellite images to identify and classify potential anomalies in the area and identify geospatial objects related to forest disaster scenarios. Utilize the available data to detect unusual patterns, segment various objects, and categorize the area to aid in effective disaster management and response planning. The data required for this task includes: (1) a main input image representing a forest area located at '/data/satellite_images/forest_area_main_image.tif', (2) a high spatial resolution (HSR) satellite image containing the geospatial area of interest located at '/data/satellite_images/geospatial_area_hsr_image.tif', (4) a main satellite image containing 10 specific spectral bands from Sentinel-2 located at '/data/satellite_images/sentinel2_spectral_bands.tif', and (5) a binary mask array specifying visible regions during the inference process located at '/data/masks/visible_regions_mask.npy'.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_main_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_hsr_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_spectral_bands.tif", "mask_path": "/data/masks/visible_regions_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 79, "task_desc": "Analyze the provided high-resolution satellite image to identify and classify various geospatial objects and anomalies related to potential natural disasters. This includes detecting landslide-prone areas in mountainous regions, identifying anomalies in forest environments that could indicate wildfire risks, and segmenting objects such as vehicles and buildings that may be affected by these disasters. The data paths for the required images and preprocessing requirements are as follows: - High-resolution satellite image for landslide detection: '/data/satellite_images/mountainous_terrain_image.jpg'- High-resolution satellite image for forest anomaly detection: '/data/satellite_images/forest_area_image.jpg'- High-resolution satellite image for geospatial object segmentation: '/data/satellite_images/geospatial_area_image.jpg'", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_image.jpg"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 80, "task_desc": "Can you analyze the high-resolution satellite image located at '/local/data/high_resolution_image.tif' to identify potential landslide-prone areas? Additionally, identify any other significant geospatial objects present in the image and provide a detailed interpretation of the identified regions. The image is a high-resolution satellite capture of a mountainous terrain. The user query describing the task is located at '/local/data/user_query.txt'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/high_resolution_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/high_resolution_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 81, "task_desc": "In the aftermath of a severe storm, assess the impact on a densely populated urban area by analyzing satellite images. You need to restore the clarity of images affected by adverse weather conditions such as rain and haze. Then, detect and count the number of individuals in crowded areas to evaluate the need for emergency response and resource allocation. Additionally, identify any objects or obstacles that may hinder rescue operations in low-light conditions. Please provide the path to the satellite images for processing. The satellite image is stored at the following paths: '/data/satellite_images/weather_degraded_image_01.jpg'. These images are affected by weather conditions and need restoration.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/weather_degraded_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 82, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Identify and classify any changes in the urban landscape, then Identify the anomalies in the urban landscape such as damaged buildings or altered infrastructure, to support disaster response and recovery efforts. Use the following data paths for the images: '/local/data/pre_disaster_image.tif', '/local/data/post_disaster_image.tif'. The pre-disaster image provides a baseline of the urban area before the event, while the post-disaster image captures the changes and damages caused by the disaster.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/local/data/pre_disaster_image.tif", "image_2_path": "/local/data/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Building_damage_assessment", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/local/data/pre_disaster_image.tif", "post_disaster_image_path": "/local/data/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 83, "task_desc": "Analyze high-resolution satellite images to identify and highlight anomalies in forest environments, such as man-made structures, water bodies, vehicles, and signs of forest health issues. Additionally, segment various geospatial objects like ships, airplanes, and vehicles within the imagery. Ensure to detect any landslide-prone or landslide-affected areas by identifying deviations from typical mountain terrain patterns. Provide the path to the satellite images for processing. The satellite images are stored in the following paths: '/data/satellite_images/forest_area_image.tif' for forest anomaly detection, '/data/satellite_images/geospatial_area_image.tif' for geospatial object segmentation, and '/data/satellite_images/mountainous_terrain_image.tif' for landslide segmentation.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 84, "task_desc": "In the aftermath of a natural disaster, assess the impact on urban areas by analyzing satellite imagery. Use high-resolution satellite images to detect objects and count crowds in affected regions, even under low-light and adverse weather conditions. This will help in understanding the extent of damage and the number of people present, aiding in efficient disaster response and resource allocation. The data required for this task includes: 1) High-resolution satellite images captured under low-light conditions, stored at 'data/satellite_images/low_light_image_01.jpg'. 2) High-resolution satellite images containing crowd scenes under adverse weather conditions, stored at 'data/satellite_images/crowd_scene_image_01.jpg'.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/crowd_scene_image_01.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 85, "task_desc": "Analyze the provided high-resolution satellite image to identify and highlight any anomalies in forest environments, such as man-made structures, water bodies, vehicles, or signs of forest health issues. Additionally, classify the image to determine the most likely categories or classes associated with the observed patterns. If the image is related to a landslide scenario, detect and highlight landslide-prone or affected areas. Please provide the path to the image data for processing. The image data can be found at the following paths: '/data/satellite_images/forest_area_image.tif' for the main input image representing a forest area, '/data/satellite_images/sentinel2_image.tif' for the main satellite image containing 10 specific spectral bands from Sentinel-2, and '/data/satellite_images/mask_array.npy' for the binary mask array specifying visible regions. For landslide analysis, use '/data/satellite_images/mountainous_terrain_image.tif' for the high-resolution aerial or satellite image capturing mountainous terrain.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_image.tif", "mask_path": "/data/satellite_images/mask_array.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 86, "task_desc": "In the aftermath of a natural disaster, analyze high-resolution satellite imagery to identify and classify affected areas and objects. Use the imagery to generate detailed descriptions of the scene, including the types of damage and the presence of critical infrastructure. This information will aid in assessing the extent of the disaster and planning effective response strategies. Please provide the path to the satellite imagery data for processing. The satellite imagery data is located at '/data/satellite_images/high_res_image.tif'. The main satellite image with specific spectral bands is located at '/data/satellite_images/spectral_bands_image.tif'. The binary mask array is available at '/data/masks/binary_mask.npy'. The user query for processing is stored at '/data/user_queries/query.txt'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/high_res_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-", "mask_path": "/data/masks/binary_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["reconstructed_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 87, "task_desc": "In the aftermath of a forest disaster, such as a forest fire, analyze high-resolution satellite images taken before and after the event to identify and map anomalies in the forest environment. This includes detecting man-made structures, water bodies, vehicles, and signs of forest health issues. Additionally, generate a detailed change map to highlight significant alterations in the landscape or changes in infrastructure. Provide a comprehensive contextual description of the detected anomalies and changes to support disaster management efforts. The data paths for the images are as follows: 'data/satellite_images/pre_disaster_image.tif' for the image taken before the disaster, 'data/satellite_images/post_disaster_image.tif' for the image taken after the disaster, and 'data/user_queries/query.txt' for the user query describing the task.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/pre_disaster_image.tif", "image_2_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 88, "task_desc": "Analyze satellite imagery to detect and classify anomalies in forest environments that could indicate potential disaster risks, such as wildfires or deforestation. The satellite image is in low-resolution. I need you to process it into high-resolution. Utilize high-resolution images to identify unusual patterns or objects, and generate an anomaly map to highlight areas of concern. Additionally, classify the detected anomalies to understand the nature of the potential threats. Please provide the path to the satellite image data for processing. The low-resolution satellite image is located at '/data/satellite_images/low_res_image.tif'. The metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata.json'. The binary mask array specifying visible regions during inference is located at '/data/satellite_images/mask_array.npy'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image.tif", "metadata_path": "/data/satellite_images/metadata.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-", "mask_path": "/data/satellite_images/mask_array.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 89, "task_desc": "In the aftermath of a severe storm, assess the impact on a densely populated urban area by analyzing images captured during the event. The images are affected by adverse weather conditions such as rain and fog, which obscure visibility. Enhance these images to restore clarity and detail, then perform object detection to identify any potential hazards or obstructions on the streets. Additionally, estimate the crowd density in public spaces to ensure the safety and efficient management of evacuation or relief efforts. Please provide the path to the folder containing the weather-degraded images for processing. The weather-degraded images are located at '/data/storm_impact/images/weather_degraded_image_01.jpg'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/storm_impact/images/weather_degraded_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 90, "task_desc": "Assess the damage to buildings in a city after a natural disaster using pre- and post-disaster satellite images. Provide a detailed description of the damage levels and identify any significant changes in infrastructure or land use. The pre-disaster satellite image is located at '/data/satellite_images/pre_disaster_image.tif', and the post-disaster satellite image is located at '/data/satellite_images/post_disaster_image.tif'. User query is stored at '/data/query/user_queries.txt'", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/query/user_queries.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 91, "task_desc": "Assess the damage caused by a recent disaster in an urban area using high-resolution satellite images taken before and after the event. Generate a detailed map indicating the level of damage to buildings and provide a comprehensive description of the affected scene. Additionally, identify any anomalies in the urban landscape that may have resulted from the disaster. Please ensure to include the path to the pre-disaster and post-disaster images, as well as any relevant metadata. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif', and the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif'. The metadata for these images can be found at '/data/satellite_images/metadata.json'. User queries are located at '/data/user_queries/query.txt'", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 92, "task_desc": "Assess the damage caused by a recent disaster in an urban area using pre- and post-disaster high-resolution satellite images. Generate a detailed damage classification map to identify the extent of damage to buildings and infrastructure. Additionally, provide a comprehensive analysis of any anomalies detected in the urban environment, and offer a contextual description of the affected areas to aid in disaster response and recovery efforts. Ensure to include the path to the pre- and post-disaster images for processing. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif' and the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif'. The user query file is located at '/data/user_queries/disaster_response_query.txt'.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/disaster_response_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 93, "task_desc": "Analyze the satellite images to identify and categorize the visible features and objects. Use this information to generate a detailed description of the scene, focusing on any potential damage or changes that may have occurred due to a recent disaster. The main input satellite image is located at '/data/satellite_images/area_of_interest_image.tif'. The user queries for risk assessment is located at '/data/risk_assessment/queries.json'.", "structured_plan": [{"agent": "RGB_GeoImage_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/area_of_interest_image.tif"}, "outputs": ["predicted_category_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_category_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/risk_assessment/queries.json"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 94, "task_desc": "Analyze high-resolution satellite images to identify and segment various geospatial objects such as ships, airplanes, and vehicles, which could be crucial for assessing the impact of a disaster event. Additionally, classify the scenes within the images to understand the types of objects or activities present. For urban areas affected by disasters, detect anomalies that deviate from typical urban patterns to help localize and analyze irregular or unexpected features. Please provide the path to the satellite image data for processing. The data paths are as follows: - High spatial resolution (HSR) satellite image: '/data/satellite_images/hsr_image.tif'- Main satellite image with spectral bands: '/data/satellite_images/sentinel2_bands.tif'- Binary mask array: '/data/satellite_images/mask_array.npy'- High-resolution satellite image of urban area: '/data/satellite_images/urban_area_image.tif'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/hsr_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_bands.tif", "mask_path": "/data/satellite_images/mask_array.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 95, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. Utilize high-resolution satellite images to restore clarity in weather-degraded visuals, enabling counting of individuals in crowded areas and accurate detection of objects in adverse weather. This will help in evaluating the extent of damage and planning effective relief operations. Please provide the path to the satellite images for processing. The satellite images are stored in the following path: '/data/satellite_images/weather_degraded_image_1.jpg'. These images are high-resolution captures of the affected urban area, showing various weather conditions such as haze, rain, and fog.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/weather_degraded_image_1.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 96, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or earthquake, utilize satellite imagery to generate high-resolution images from available low-resolution multi-spectral data. These high-resolution images will then be analyzed to classify affected areas and segment various geospatial objects. This information will aid in disaster response and recovery efforts by providing detailed insights into the affected areas. The low-resolution multi-spectral image data can be found at '/data/satellite_images/low_res/multi_spectral_image.tif'. The metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata/contextual_metadata.json'. The binary mask array specifying visible regions is located at '/data/satellite_images/masks/visibility_mask.npy'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res/multi_spectral_image.tif", "metadata_path": "/data/satellite_images/metadata/contextual_metadata.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-", "mask_path": "/data/satellite_images/masks/visibility_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 97, "task_desc": "Assess the damage to buildings caused by a recent disaster by analyzing pre- and post-disaster satellite images but some of them are in low-resolution. Utilize high-resolution images to enhance the accuracy of the damage classification. Additionally, classify the temporal sequence of satellite images to understand the progression of the disaster's impact over time. Ensure that the analysis incorporates both spatial and temporal features to provide a comprehensive assessment of the disaster's effects. The data paths are as follows: 'data/satellite_images/low_res_image_T.jpg' for low-resolution satellite image at time T, 'data/satellite_images/high_res_image_T.jpg' for high-resolution satellite image at time T, 'data/satellite_images/low_res_image_T_prime.jpg' for low-resolution satellite image at time T', 'data/metadata/metadata.json' for metadata associated with the satellite images, 'data/satellite_images/pre_disaster_image.jpg' for the satellite image of the area of interest taken before the disaster, 'data/satellite_images/post_disaster_image.jpg' for the satellite image of the area of interest taken after the disaster, and 'data/satellite_images/image_sequence/temporal_sequence.npy' for a temporal sequence of satellite images.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/satellite_images/low_res_image_T.jpg", "high_res_image_T_path": "data/satellite_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "data/satellite_images/low_res_image_T_prime.jpg", "metadata_path": "data/metadata/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"pre_disaster_image_path": "data/satellite_images/pre_disaster_image.jpg", "post_disaster_image_path": "-0-"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "data/satellite_images/image_sequence/temporal_sequence.npy"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 98, "task_desc": "Analyze the impact of a recent disaster in an urban area by identifying anomalies in high-resolution satellite images taken before and after the event. Use the anomaly map to generate a detailed report on the affected regions, highlighting any significant deviations from normal urban patterns. Additionally, classify the type of disaster based on visual features extracted from the images and provide a comprehensive overview of the situation, including potential risks and areas requiring immediate attention. Ensure to include the path to the high-resolution images in the analysis. The high-resolution satellite images before the disaster are located at '/data/satellite_images/before_disaster/high_res_image_before.png' and the images after the disaster are located at '/data/satellite_images/after_disaster/high_res_image_after.png'. The user query describing the task is located at '/data/user_queries/disaster_analysis_query.txt'", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/after_disaster/high_res_image_after.png"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/disaster_analysis_query.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "RGB_GeoImage_Classifier", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["predicted_category_path"]}], "source": "benchmark"} {"task_id": 99, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Identify and categorize any anomalies in urban areas and then figure out damages to buildings and other structures. Additionally, classify the land use and any changes in the urban landscape that may have occurred as a result of the disaster. Please provide the path to the pre- and post-disaster images for analysis. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image_urban_area.tif' and the post-disaster image is located at '/data/satellite_images/post_disaster_image_urban_area.tif'.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image_urban_area.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image_urban_area.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image_urban_area.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "RGB_GeoImage_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image_urban_area.tif"}, "outputs": ["predicted_category_path"]}], "source": "benchmark"} {"task_id": 100, "task_desc": "Analyze the impact of a recent disaster on an urban area by detecting changes in infrastructure and identifying anomalies in the affected region. Use high-resolution satellite images taken before and after the disaster to generate a change map and an anomaly map. Provide a detailed contextual description of the anomalies observed, focusing on newly constructed buildings, damaged structures, and any unexpected features that deviate from typical urban patterns. Ensure the analysis supports decision-making for urban recovery and planning efforts. The data paths for the images are as follows: 'data/satellite_images/pre_disaster_image.tif' for the image taken before the disaster, 'data/satellite_images/post_disaster_image.tif' for the image taken after the disaster, and 'data/satellite_images/urban_area_image.tif' for the high-resolution image of the urban area. The user queries are located at 'data/user_queries/task_description.txt'", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/pre_disaster_image.tif", "image_2_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "data/user_queries/task_description.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 101, "task_desc": "In the event of a flood, there is predicted the water depths across a region and assess the impact on traffic speed. Utilize available precipitation data to forecast future rainfall patterns, convert this data into a suitable format for flood depth prediction, and then determine how the predicted water depths affect traffic flow on road networks. Ensure all necessary data files are correctly referenced in the process. The data paths are as follows: 'radar_echo_frames_path': '/local/data/radar_echo_frames/input_frames.dat', 'region_topography_path': '/local/data/topography/region_dem.tif', 'precipitation_data_path': '/local/data/precipitation/precipitation_timeseries.csv', 'road_network_path': '/local/data/road_network/road_network_data.shp', 'flood_depth_path': '/local/data/flood_depth/flood_depth_data.csv'.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/local/data/radar_echo_frames/input_frames.dat"}, "outputs": ["predicted_radar_echo_frames_path"]}, {"agent": "precipitation_data_convert_tool", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_radar_echo_frames_path"]}, "inputs": {"radar_echo_frames_path": "-0-"}, "outputs": ["time_series_array_path"]}, {"agent": "Flood_depth_prediction", "step": 2, "dependence": [1], "dependence_content": {"1": ["time_series_array_path"]}, "inputs": {"region_topography_path": "/local/data/topography/region_dem.tif", "precipitation_data_path": "-1-"}, "outputs": ["predicted_water_depths_path"]}, {"agent": "depth_speed_model", "step": 3, "dependence": [2], "dependence_content": {"2": ["predicted_water_depths_path"]}, "inputs": {"road_network_path": "/local/data/road_network/road_network_data.shp", "flood_depth_path": "-2-"}, "outputs": ["traffic_speed_path"]}], "source": "benchmark"} {"task_id": 102, "task_desc": "Analyze the provided high-resolution satellite image to identify and classify various geospatial objects, such as buildings, vehicles, and natural features. Additionally, detect any anomalies in urban areas that may indicate disaster events, such as structural damage or unusual patterns. Provide a detailed report on the identified objects and anomalies, including their locations and potential implications for disaster management. The high-resolution satellite image is located at '/data/satellite_images/urban_area_high_res.tif'. The user query describing the task is located at '/data/user_queries/query.txt'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_high_res.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 103, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. Utilize high-resolution satellite imagery to detect and count the number of people in crowded areas, even under challenging weather conditions like fog or rain. Additionally, identify and describe any significant objects or structures that may have been affected by the storm. This information will be crucial for coordinating emergency response efforts and ensuring the safety of the affected population. The satellite image data can be found at '/data/satellite_images/storm_aftermath_image_01.tif'. The user query describing the task is located at '/data/user_queries/storm_impact_query.txt'.", "structured_plan": [{"agent": "Crowd_Counting_in_Adverse_Weather", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/storm_aftermath_image_01.tif"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/storm_aftermath_image_01.tif"}, "outputs": ["detected_objects_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/storm_aftermath_image_01.tif", "user_query_path": "/data/user_queries/storm_impact_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 104, "task_desc": "In the aftermath of a natural disaster, assess the impact on a forested area by analyzing high-resolution satellite images taken after the event. Identify and map any anomalies such as forest fires or other disruptions to the forest environment. Additionally, evaluate the potential for landslides in mountainous regions affected by the disaster. Use the temporal sequence of images to classify the changes over time and generate insights for disaster management and recovery efforts. The data paths are as follows: Post-disaster forest image: '/data/satellite_images/post_disaster_forest_image.tif' Post-disaster mountainous terrain image: '/data/satellite_images/post_disaster_mountain_image.tif' Temporal sequence of images: '/data/satellite_images/temporal_sequence_image_01.tif', '/data/satellite_images/temporal_sequence_image_02.tif', '/data/satellite_images/temporal_sequence_image_03.tif'.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_forest_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_mountain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence_image_01.tif"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence_image_02.tif"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 4, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence_image_03.tif"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 105, "task_desc": "We need to assess the impact on infrastructure and natural landscapes. Utilize high-resolution satellite imagery to detect and identify objects such as damaged buildings, blocked roads, and affected vegetation. Ensure that the images are clear and free from weather-related distortions to improve the accuracy of object detection. Additionally, provide a detailed contextual description of the detected objects to aid in disaster response and recovery efforts. Please use the following paths for the initial satellite imagery and related data: /local/data/satellite/image1.tif, /local/data/user_query.txt. The satellite images are high-resolution TIFF files capturing the affected areas, and the user query file contains specific instructions for the analysis.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/satellite/image1.tif"}, "outputs": ["restored_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/local/data/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 106, "task_desc": "In the aftermath of a severe storm, assess the impact on a densely populated urban area by generating high-resolution satellite images based on the following description: the densely populated urban area is affected by the snowstorm seeking help. Use these images to estimate the crowd density and count in various locations, ensuring accurate analysis despite adverse weather conditions. Additionally, restore any weather-degraded images to enhance clarity and detail, facilitating better decision-making for emergency response and resource allocation. The data paths for this task are as follows: 'data/captions/storm_description.txt' for the descriptive caption, 'data/metadata/storm_metadata.json' for the metadata.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/captions/storm_description.txt", "metadata_path": "data/metadata/storm_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 107, "task_desc": "I need to assess the impact of recent natural disasters on urban infrastructure. Could you analyze the satellite images taken before and after the event to identify any significant changes or damages? Please provide a detailed description of the affected areas and any notable alterations in the landscape. The images are stored in the following paths: '/local/data/satellite_images/before_event_image_1.tif' and '/local/data/satellite_images/after_event_image_2.tif'. Query file is located at '/local/data/user_queries/urban_infrastructure_impact_query.txt'", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/local/data/satellite_images/before_event_image_1.tif", "image_2_path": "/local/data/satellite_images/after_event_image_2.tif"}, "outputs": ["change_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/urban_infrastructure_impact_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 108, "task_desc": "In the aftermath of a natural disaster, we need to assess the impact on the landscape and infrastructure. Using high-resolution satellite images taken before and after the event, identify and map any significant changes in the area, such as new construction or damage to existing structures. Additionally, detect and highlight any landslide-prone or affected regions to prioritize areas for immediate intervention. Please provide the path to the high-resolution images for analysis. The paths to the images are as follows: 'data/images/before_event_satellite_image.tif' for the image taken before the event, 'data/images/after_event_satellite_image.tif' for the image taken after the event, and 'data/images/mountainous_terrain_image.tif' for the image capturing the mountainous terrain.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/images/before_event_satellite_image.tif", "image_2_path": "data/images/after_event_satellite_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 109, "task_desc": "Analyze the provided forest imagery to identify and highlight any anomalies such as man-made structures, water bodies, vehicles, or signs of forest health issues. Generate an anomaly map to visualize these deviations. Additionally, provide a comprehensive scene interpretation and contextual description of the detected anomalies to assist in disaster management efforts, such as assessing potential risks or planning mitigation strategies. The main input image is located at '/data/forest_images/forest_area_image.jpg'. The user query describing the task is located at '/data/user_queries/forest_anomaly_detection_query.txt'.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/forest_images/forest_area_image.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/forest_anomaly_detection_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 110, "task_desc": "Analyze the provided high-resolution satellite image to identify and highlight any anomalies or potential disaster-related features in a forest environment. Use the image to detect deviations from typical forest patterns, such as signs of wildfires, diseased trees, or unauthorized constructions. Provide a detailed contextual description of the identified anomalies to assist in disaster management efforts. Ensure the image is located at the specified path: '/local/data/forest_image/high_res_forest_image.tif'. The image is a high-resolution TIFF file capturing a forest area, suitable for detailed analysis. The analysis queries are located at '/local/data/user_queries/forest_analysis_query.txt'", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/forest_image/high_res_forest_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/forest_analysis_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 111, "task_desc": "Analyze high-resolution satellite images to identify and highlight anomalies in urban areas affected by disaster events. Additionally, segment and classify various geospatial objects within these images to assist in disaster response and recovery efforts. Please provide the path to the satellite images for processing. The satellite images are stored at '/data/satellite_images/urban_disaster_event_01.tif'. ", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_disaster_event_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_disaster_event_01.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 112, "task_desc": "Assess the impact of a recent disaster on a forested area by analyzing high-resolution satellite images taken before and after the event. Identify and map any anomalies such as man-made structures, water bodies, vehicles, or signs of forest health issues. Additionally, evaluate the extent of damage to buildings within the affected area, classifying them into categories such as no damage, minor damage, major damage, or destroyed. Ensure the images used are related to forest scenarios, and the disaster type could include events like forest fires. Please provide the path to the high-resolution images for processing. The paths to the required images are as follows: 'data/images/forest_area_main_input.jpg' for the main input image representing a forest area, 'data/images/pre_disaster_forest.jpg' for the satellite image taken before the disaster, and 'data/images/post_disaster_forest.jpg' for the satellite image taken after the disaster.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/images/forest_area_main_input.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "data/images/pre_disaster_forest.jpg", "post_disaster_image_path": "data/images/post_disaster_forest.jpg"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 113, "task_desc": "In the aftermath of a severe storm, assess the impact on urban infrastructure by analyzing satellite images. First, restore the clarity of images affected by adverse weather conditions such as rain or haze. Then, enhance the visibility of these images to detect and identify objects like vehicles, buildings, and debris in dark night conditions. This information will aid in evaluating the extent of damage and prioritizing emergency response efforts. Please provide the path to the satellite images for processing. The satellite images are stored at the following paths: '/data/satellite_images/weather_degraded_image_01.jpg', and '/data/satellite_images/low_light_image_01.jpg'. These images include those affected by weather conditions and those captured under low-light conditions.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/weather_degraded_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 114, "task_desc": "Identify and map potential landslide-prone areas using high-resolution satellite imagery. Additionally, detect and highlight any anomalies in forested regions that could indicate potential disasters such as wildfires. The high-resolution satellite imagery data for landslide detection is located at '/data/satellite_images/landslide_detection/high_res_image_01.tif'. The forest imagery data for anomaly detection is located at '/data/satellite_images/forest_anomalies/forest_image_01.tif'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/landslide_detection/high_res_image_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_anomalies/forest_image_01.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 115, "task_desc": "Assess the damage to buildings after a disaster by generating high-resolution images of the affected area using available low-resolution and high-resolution images taken before the disaster, along with low-resolution images taken after the disaster. Use these generated high-resolution images to classify the extent of damage to buildings into categories such as no damage, minor damage, major damage, or destroyed. Ensure that the spatial and temporal consistency of the images is maintained throughout the process. Please provide the path to the pre-disaster low-resolution image, pre-disaster high-resolution image, and post-disaster low-resolution image. The paths to the required data files are as follows: Pre-disaster low-resolution image: '/data/pre_disaster/low_res_image_T.jpg', Pre-disaster high-resolution image: '/data/pre_disaster/high_res_image_T.jpg', Post-disaster low-resolution image: '/data/post_disaster/low_res_image_T_prime.jpg', Metadata: '/data/metadata/satellite_metadata.json'.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/data/pre_disaster/low_res_image_T.jpg", "high_res_image_T_path": "/data/pre_disaster/high_res_image_T.jpg", "low_res_image_T_prime_path": "/data/post_disaster/low_res_image_T_prime.jpg", "metadata_path": "/data/metadata/satellite_metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"pre_disaster_image_path": "/data/pre_disaster/high_res_image_T.jpg", "post_disaster_image_path": "-0-"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 116, "task_desc": "Generate a high-resolution satellite image to assess the impact of a recent natural disaster. Use available metadata and textual descriptions to create an initial low-resolution multi-spectral image of the affected area. Then, enhance this image to a high-resolution format to better analyze the extent of the damage and aid in disaster response efforts. Ensure the data path for the metadata and textual descriptions is correctly set. The metadata and textual descriptions are stored in the following paths: '/data/metadata/metadata_vector.json' for metadata and '/data/textual_descriptions/description.txt' for textual descriptions.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/data/textual_descriptions/description.txt", "metadata_path": "/data/metadata/metadata_vector.json"}, "outputs": ["generated_image_path"]}, {"agent": "High-Resolution_Image_Reconstructor", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-", "metadata_path": "/data/metadata/metadata_vector.json"}, "outputs": ["high_resolution_image_path"]}], "source": "benchmark"} {"task_id": 117, "task_desc": "Analyze high-resolution satellite images to identify and highlight anomalies in urban areas that may indicate disaster-related events. Use the anomaly map to detect irregular patterns or unexpected features that could signify damage or disruption. Additionally, classify the multi-spectral data from the satellite images to categorize and understand the affected regions, providing insights into the types of structures or landscapes impacted by the disaster. Ensure the data path for the satellite images is correctly set for processing. The high-resolution satellite image of an urban area is located at '/data/satellite_images/urban_area_high_res.tif'. The main satellite image containing 10 specific spectral bands from Sentinel-2 is located at '/data/satellite_images/sentinel2_bands.tif'. The binary mask array specifying visible regions is located at '/data/masks/urban_area_mask.npy'.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_high_res.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_bands.tif", "mask_path": "/data/masks/urban_area_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 118, "task_desc": "Can you analyze the satellite imagery to identify and classify any potential disaster-related anomalies in both forest and urban environments? The satellite imagery data is located at the specified paths: '/data/satellite_images/forest_area_image_2023.tif' for forest areas and '/data/satellite_images/urban_area_image_2023.tif' for urban areas. The forest area image is a high-resolution TIFF file capturing a dense forest region, while the urban area image is a high-resolution TIFF file capturing a metropolitan region.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image_2023.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image_2023.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 119, "task_desc": "In the aftermath of a forest disaster, such as a fire, analyze high-resolution satellite images to identify and classify changes and anomalies in the forest environment. Use the provided high-resolution images taken before and after the disaster to detect any unusual patterns or structures, such as damaged areas, new water bodies, or man-made structures. Additionally, classify the visible features in the post-disaster image to understand the impact and categorize the changes observed. Ensure the images are related to forest scenarios and provide the path to the images for processing. The paths to the images are as follows: 'data/forest_disaster/pre_disaster_image.tif' for the image taken before the disaster and 'data/forest_disaster/post_disaster_image.tif' for the image taken after the disaster.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/forest_disaster/pre_disaster_image.tif", "image_2_path": "data/forest_disaster/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "RGB_GeoImage_Classifier", "step": 1, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["predicted_category_path"]}], "source": "benchmark"} {"task_id": 120, "task_desc": "Predict the future precipitation patterns to aid in flood risk assessment and disaster management. Utilize the radar echo data available at '/local/data/radar_echo_data/radar_echo_20231001_0000.nc' to forecast the intensity and distribution of rainfall over the next 100 minutes. Convert the predicted radar echo frames into a time-series format suitable for flood depth prediction models, ensuring the data is log-transformed and normalized as per the training data requirements. The radar echo data file 'radar_echo_20231001_0000.nc' contains observed precipitation data over time, which will be used as input for the prediction model.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/local/data/radar_echo_data/radar_echo_20231001_0000.nc"}, "outputs": ["predicted_radar_echo_frames_path"]}, {"agent": "precipitation_data_convert_tool", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_radar_echo_frames_path"]}, "inputs": {"radar_echo_frames_path": "-0-"}, "outputs": ["time_series_array_path"]}], "source": "benchmark"} {"task_id": 121, "task_desc": "In the aftermath of a natural disaster, it's crucial to assess the impact on forested areas. Using high-resolution satellite images taken before and after the disaster, identify and map the changes in the landscape. Additionally, detect any anomalies in the forest environment, such as new structures or signs of forest health issues, to aid in recovery and management efforts. Please provide the path to the pre-disaster and post-disaster images for analysis. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif' and the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif'.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/data/satellite_images/pre_disaster_image.tif", "image_2_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 122, "task_desc": "In the aftermath of a severe storm, we need to assess the crowd density in a heavily affected urban area to coordinate emergency response efforts. The storm has caused significant weather-related degradation in the surveillance images we have. Please process these images to restore their clarity and then estimate the crowd count to help us allocate resources effectively. The images are located at the path: '/local/data/weather_degraded_images/image_001.jpg'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/weather_degraded_images/image_001.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 123, "task_desc": "Assess the damage caused by a natural disaster in an urban area by analyzing pre- and post-disaster high-resolution satellite images. Additionally, identify any anomalies in the urban landscape that may have resulted from the disaster. Ensure to use the appropriate data paths for the images required for this analysis. The pre-disaster image is stored at '/data/satellite_images/pre_disaster_image_urban_area.tif', and the post-disaster image is stored at '/data/satellite_images/post_disaster_image_urban_area.tif'. These images are high-resolution satellite captures of the urban area before and after the disaster, respectively.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image_urban_area.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image_urban_area.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 124, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or flood, it is crucial to assess the damage and identify areas that require immediate attention. Using high-resolution satellite imagery captured during foggy conditions, analyze the images to detect and identify objects such as damaged infrastructure, vehicles, and debris. Provide a detailed contextual description of the scene to aid in disaster management and response efforts. The path to the satellite imagery is: /local/data/satellite_images/foggy_conditions_image1.jpg. The path to the user query is: /local/data/user_queries/disaster_management_query.txt.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/foggy_conditions_image1.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/disaster_management_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 125, "task_desc": "In the aftermath of a natural disaster, assess the extent of damage to buildings in a city using satellite imagery. Utilize pre- and post-disaster images to classify the damage levels of buildings into categories such as no damage, minor damage, major damage, and destroyed. Additionally, provide a detailed contextual description of the affected areas to aid in disaster response and recovery efforts. Please ensure to include the path to the pre-disaster image, post-disaster image, and any relevant metadata. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image_2023.tif', the post-disaster image is located at '/data/satellite_images/post_disaster_image_2023.tif', and the metadata is located at '/data/metadata/disaster_metadata_2023.json'. User query at '/data/user_queries.txt'", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image_2023.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image_2023.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 126, "task_desc": "Can you analyze the satellite images taken before and after a recent disaster to assess the extent of building damage? I need a detailed classification of the damage levels, such as no damage, minor damage, major damage, or destroyed. Additionally, provide a comprehensive description of the affected area, highlighting any significant changes or anomalies detected in the forest regions. Please ensure the analysis is thorough and includes any relevant contextual information. The pre-disaster satellite image is located at '/data/satellite_images/pre_disaster_image.jpg' and the post-disaster satellite image is located at '/data/satellite_images/post_disaster_image.jpg'. The user query describing the task is located at '/data/user_queries/disaster_analysis_query.txt'.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.jpg", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.jpg"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/disaster_analysis_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 127, "task_desc": "Analyze a high-resolution satellite image to identify and classify various objects or scenes within the image. Additionally, detect and highlight any anomalies in urban areas that may indicate disaster events, such as landslides or other disruptions. Provide a detailed classification of the identified objects and an anomaly map to assist in disaster management efforts. Please ensure the image data is accessible and specify the path to the image file. The main satellite image containing 10 specific spectral bands from Sentinel-2 (excluding B1, B9, and B10) is located at '/local/data/satellite_images/sentinel2_image.tif'. The binary mask array specifying which regions of the satellite image are visible (value 0) or masked (value 1) during the inference process is located at '/local/data/masks/visibility_mask.npy'. The high-resolution satellite image of an urban area is located at '/local/data/satellite_images/urban_area_image.tif'.", "structured_plan": [{"agent": "Multi_Spectral_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/sentinel2_image.tif", "mask_path": "/local/data/masks/visibility_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 128, "task_desc": "In the aftermath of a recent disaster, we need to assess the extent of landslide damage in a mountainous region. We have access to high-resolution satellite images taken before and after the disaster. Please analyze these images to identify and highlight areas that are prone to or affected by landslides. Additionally, classify the sequence of images to understand the temporal changes and categorize the affected areas based on their visual features. This will help in planning effective disaster response and management strategies. Please ensure the images are related to landslides and provide the path to the images for processing. The paths to the images are as follows: '/local/data/landslide_images/before_disaster_image_01.tif', '/local/data/landslide_images/after_disaster_image_01.tif', '/local/data/landslide_images/temporal_sequence_image_01.tif', '/local/data/landslide_images/temporal_sequence_image_02.tif'. The images are high-resolution satellite captures of the mountainous terrain, with temporal sequences showing changes over time.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/landslide_images/before_disaster_image_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/landslide_images/after_disaster_image_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/landslide_images/temporal_sequence_image_01.tif"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/landslide_images/temporal_sequence_image_02.tif"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 129, "task_desc": "Analyze high-resolution satellite images to identify and segment various geospatial objects such as ships, airplanes, and vehicles. Additionally, classify sequences of these images to understand the temporal changes and categorize the affected areas based on their spatial and temporal features. Ensure to use the high-resolution images captured before and after the disaster for accurate analysis. The high-resolution satellite images are stored in '/data/satellite_images/high_res_image_01.tif' and '/data/satellite_images/high_res_image_02.tif'. The temporal sequence of images is stored in '/data/satellite_images/temporal_sequence_01.tif' and '/data/satellite_images/temporal_sequence_02.tif'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/high_res_image_01.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/high_res_image_02.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence_01.tif"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence_02.tif"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 130, "task_desc": "Analyze the satellite image to identify and categorize visible objects or features that could be impacted by a natural disaster, such as buildings, infrastructure, or land use types. Provide a detailed contextual description of the scene, highlighting potential vulnerabilities and areas of concern for disaster management efforts. Please ensure the image is located at the specified path: '/local/data/satellite_images/disaster_area_01.tif'. The user query describing the task is located at '/local/data/user_queries/disaster_analysis_query.txt'.", "structured_plan": [{"agent": "RGB_GeoImage_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/disaster_area_01.tif"}, "outputs": ["predicted_category_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_category_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/disaster_analysis_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 131, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. Please analyze the high-resolution satellite images to count the number of people present in the affected zones. Additionally, identify any objects or structures that may have been damaged or are at risk due to the storm. The data required for this task is available locally at the specified paths: '/local/data/satellite_images/crowd_scene_image.jpg' for the primary input image containing a crowd scene, and '/local/data/satellite_images/low_light_image.jpg' for the input image captured under low-light conditions. These images are high-resolution satellite captures of the affected urban area, taken shortly after the storm.", "structured_plan": [{"agent": "Crowd_Counting_in_Adverse_Weather", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/crowd_scene_image.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/low_light_image.jpg"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 132, "task_desc": "Predict the potential flood water depths over a region by analyzing radar echo frames and precipitation data. Use the radar data to convert it into a suitable format for precipitation analysis, and then predict the water depths dynamically considering the spatial and temporal variability in rainfall. Ensure that the data is log-transformed and normalized as per the training data requirements. Please provide the path to the radar echo frames data for processing. The radar echo frames data is located at '/local/data/radar_echo_frames/radar_echo_2023_10_01.dat'. The region topography data is located at '/local/data/topography/region_dem_2023_10_01.tif'.", "structured_plan": [{"agent": "precipitation_data_convert_tool", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/local/data/radar_echo_frames/radar_echo_2023_10_01.dat"}, "outputs": ["time_series_array_path"]}, {"agent": "Flood_depth_prediction", "step": 1, "dependence": [0], "dependence_content": {"0": ["time_series_array_path"]}, "inputs": {"region_topography_path": "/local/data/topography/region_dem_2023_10_01.tif", "precipitation_data_path": "-0-"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 133, "task_desc": "Assess the impact of a recent disaster on a forested area by analyzing high-resolution satellite images. Identify and map any anomalies in the area. Detect and highlight any significant changes in the landscape to support disaster management and recovery efforts. The data includes: 1) Pre-disaster satellite image: 'data/satellite_images/pre_disaster_image.tif'. 2) Post-disaster satellite image: 'data/satellite_images/post_disaster_image.tif'.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/pre_disaster_image.tif", "image_2_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}], "source": "benchmark"} {"task_id": 134, "task_desc": "In the aftermath of a recent natural disaster, we need to assess the extent of landslide damage in a mountainous region. Utilize high-resolution satellite imagery to identify and map landslide-prone or affected areas. Additionally, classify the types of terrain and vegetation present in the region to aid in understanding the impact and planning recovery efforts. Please provide the path to the high-resolution satellite image data for processing. The high-resolution satellite image data is located at '/data/satellite_images/mountainous_region_high_res.tif'. The main satellite image containing 10 specific spectral bands from Sentinel-2 is located at '/data/satellite_images/sentinel2_bands.tif'. The binary mask array specifying visible and masked regions is located at '/data/masks/visibility_mask.npy'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_region_high_res.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_bands.tif", "mask_path": "/data/masks/visibility_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 135, "task_desc": "Analyze the provided high-resolution satellite image to identify and classify any anomalies or unusual patterns in a forest environment, such as signs of wildfires, man-made structures, or other potential threats. Additionally, determine the specific categories or classes associated with the detected features to aid in disaster management and response planning. Please ensure the image is located at the specified path for processing. The main input image is located at '/data/satellite_images/forest_area_image.tif'. The binary mask array is located at '/data/masks/forest_area_mask.npy'.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "mask_path": "/data/masks/forest_area_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 136, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or earthquake, it is crucial to assess the impact on infrastructure and human presence in affected areas. Utilize high-resolution satellite images to detect and identify objects, such as vehicles and buildings, in low-light conditions to understand the extent of damage. Additionally, estimate the crowd density in these areas to aid in efficient resource allocation and rescue operations. The high-resolution satellite images are stored at '/data/satellite_images/low_light_image_01.jpg' for low-light object detection and '/data/satellite_images/crowd_scene_image_01.jpg' for crowd counting in adverse weather conditions.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/crowd_scene_image_01.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 137, "task_desc": "Analyze the provided high-resolution remote sensing imagery to identify and map potential landslide-prone areas. Generate a detailed contextual description of the identified regions, highlighting the degree of anomaly and potential risk factors. Use the anomaly maps to facilitate a comprehensive understanding of the affected areas for effective disaster management and intervention planning. The high-resolution imagery is stored at 'data/remote_sensing/imagery/high_res_image.tif'. The user query describing the task is located at 'data/user_queries/landslide_analysis_query.txt'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/remote_sensing/imagery/high_res_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "data/user_queries/landslide_analysis_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 138, "task_desc": "In the aftermath of a natural disaster, analyze satellite imagery to identify and classify affected areas. Use the classification results to generate a detailed report that includes contextual descriptions of the damage and affected regions. This report will aid in understanding the extent of the disaster and assist in planning relief efforts. Please provide the path to the satellite imagery data for processing. The satellite imagery data is located at '/data/satellite_images/sentinel2_image.tif', and the binary mask array is located at '/data/masks/visibility_mask.npy'. The user query describing the task is located at '/data/queries/user_query.txt'.", "structured_plan": [{"agent": "Multi_Spectral_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel2_image.tif", "mask_path": "/data/masks/visibility_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["reconstructed_image_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/queries/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 139, "task_desc": "In the aftermath of a disaster in an urban area, assess the extent of anomalies by generating a high-resolution image of the affected location using pre-disaster low and high-resolution images, and a post-disaster low-resolution image. Then, analyze the generated high-resolution image to identify and map anomalies that deviate from typical urban patterns, providing insights into the impact and irregularities caused by the disaster. Ensure the images used are related to urban scenarios and the disaster type is applicable to urban environments. Please provide the paths to the low-resolution image at time T, high-resolution image at time T, low-resolution image at time T', and metadata for processing. The paths are as follows: 'data/pre_disaster_low_res_T.jpg', 'data/pre_disaster_high_res_T.jpg', 'data/post_disaster_low_res_T_prime.jpg', and 'data/metadata.json'. The images are satellite images capturing urban areas before and after the disaster, and the metadata includes information such as latitude, longitude, ground sampling distance (GSD), cloud cover fraction, year, month, and day.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/pre_disaster_low_res_T.jpg", "high_res_image_T_path": "data/pre_disaster_high_res_T.jpg", "low_res_image_T_prime_path": "data/post_disaster_low_res_T_prime.jpg", "metadata_path": "data/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 140, "task_desc": "In the event of a natural disaster, such as a severe storm or flood, analyze high-resolution satellite images to detect objects and assess crowd density in affected areas. Utilize the available satellite imagery to identify key objects and estimate the number of people in various locations, even under challenging weather conditions like fog or rain. This information will aid in coordinating emergency response efforts and allocating resources effectively. The satellite images are stored in the following paths: '/data/satellite_images/foggy_conditions/image_foggy_01.tif' for images captured in foggy conditions and '/data/satellite_images/crowd_scenes/image_crowd_01.tif' for images containing crowd scenes.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/foggy_conditions/image_foggy_01.tif"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/crowd_scenes/image_crowd_01.tif"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 141, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by analyzing satellite images taken before and after the event. Utilize high-resolution imagery to detect changes in the landscape and classify the extent of damage to buildings. Provide a detailed map indicating areas of no damage, minor damage, major damage, and destroyed structures. Ensure the analysis incorporates both spatial and temporal information to accurately reflect the current state of the affected region. Please provide the path to the pre-disaster and post-disaster images along with any relevant metadata. The pre-disaster image is located at '/data/satellite_images/pre_disaster_image.tif' and the post-disaster image is located at '/data/satellite_images/post_disaster_image.tif'.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 142, "task_desc": "Analyze the high-resolution satellite imagery to identify and classify geospatial objects such as buildings, vehicles, and other structures. Generate a detailed contextual description of the scene, highlighting any anomalies or potential disaster-related changes, such as landslides or urban disturbances. Provide insights into the current state of the area to aid in disaster management and response planning. Please ensure to use the image located at '/local/data/high_resolution_image.jpg'. The user query is located at '/local/data/user_query.txt'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/high_resolution_image.jpg"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 143, "task_desc": "In the aftermath of a natural disaster, assess the extent of damage and changes in urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Utilize the images to identify significant changes in the landscape, such as newly constructed or damaged buildings, and generate a detailed change map. Additionally, perform geospatial object segmentation to classify and map various objects like buildings, vehicles, and other infrastructure elements within the affected area. This comprehensive analysis will aid in effective disaster response and recovery planning. Please provide the path to the high-resolution images taken before and after the disaster. The paths to the images are as follows: 'data/satellite_images/before_disaster_image.tif' for the image taken before the disaster, and 'data/satellite_images/after_disaster_image.tif' for the image taken after the disaster.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/before_disaster_image.tif", "image_2_path": "data/satellite_images/after_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 144, "task_desc": "In the event of negative weather of foggy environments, it is crucial to assess the impact on infrastructure and the environment. Enhance the clarity of these images to ensure accurate detection and assessment of the damage. Utilize high-resolution satellite images to detect and identify objects and structures that may have been affected by the adverse weather conditions. This will aid in coordinating effective disaster response and recovery efforts. Please provide the path to the satellite images for processing. The satellite images are stored at the following paths: '/data/satellite_images/foggy_environment_image_01.jpg' and '/data/satellite_images/foggy_environment_image_02.jpg'. These images are high-resolution captures of areas affected by foggy weather conditions, intended for object detection and image restoration tasks.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/foggy_environment_image_01.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/foggy_environment_image_02.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 3, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 145, "task_desc": "Generate a high-resolution satellite image from a low-resolution multi-spectral image to identify and classify potential anomalies in a forest environment, such as wildfires or other disasters. Use the path '/local/data/low_resolution_image.tif' for the initial image data and '/local/data/metadata_vector.json' for the metadata vector providing contextual details for the satellite image.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/low_resolution_image.tif", "metadata_path": "/local/data/metadata_vector.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 146, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or earthquake, it is crucial to assess the impact on infrastructure and identify any potential hazards. Using high-resolution satellite images, analyze the affected area to detect and identify objects such as damaged buildings, vehicles, and debris. Additionally, consider the challenges posed by low-light conditions or foggy weather that may obscure visibility, and ensure that these factors are accounted for in the detection process. This analysis will aid in coordinating emergency response efforts and prioritizing areas that require immediate attention. The high-resolution satellite images are stored in the following paths: '/data/satellite_images/low_light_image_01.jpg' for low-light conditions and '/data/satellite_images/foggy_image_01.jpg' for foggy or normal weather conditions.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/foggy_image_01.jpg"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 147, "task_desc": "In the aftermath of a severe weather event, assess the impact on infrastructure and identify any potential hazards by analyzing high-resolution satellite images. Utilize advanced image restoration techniques to enhance the clarity of weather-degraded images, and perform object detection to identify critical objects or structures that may have been affected. This analysis will aid in disaster response and recovery efforts by providing detailed insights into the affected areas. The high-resolution satellite images are stored at '/data/satellite_images/weather_event_image_01.tif'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/weather_event_image_01.tif"}, "outputs": ["restored_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 148, "task_desc": "Analyze the impact of a recent disaster on an urban area by comparing high-resolution satellite images taken after the event by identifying and segmenting geospatial objects. Additionally, detect any anomalies in the urban landscape that may indicate areas of significant change or damage. Use the following data paths for the images: '/local/data/disaster_image.tif'. The images are high-resolution satellite images capturing the urban area of interest before and after the disaster, respectively.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/disaster_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 149, "task_desc": "I need to analyze satellite imagery to identify potential landslide-prone areas. The initial data is a low-resolution multi-spectral image, and I want to generate a high-resolution image from it. Once I have the high-resolution image, I need to detect any anomalies that might indicate landslide risks. Please ensure the analysis is thorough and provides a clear indication of areas that might require further investigation or intervention. The path to the low-resolution multi-spectral image is: /data/satellite_images/low_res_image.tif. The path to the metadata vector providing contextual details for the satellite image is: /data/satellite_images/metadata_vector.json.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image.tif", "metadata_path": "/data/satellite_images/metadata_vector.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 150, "task_desc": "Analyze high-resolution satellite images taken before and after a forest disaster, such as a forest fire, to identify and map anomalies. These anomalies could include changes in forest health, presence of man-made structures, or water bodies. Use the anomaly map to assess the impact of the disaster and classify the temporal sequence of images to understand the progression and effects over time. Please provide the path to the high-resolution images for processing. The data paths for the images are as follows: - Post-disaster image: '/data/satellite_images/post_disaster_image.tif'- Temporal sequence images: '/data/satellite_images/temporal_sequence/image_01.tif', '/data/satellite_images/temporal_sequence/image_02.tif', '/data/satellite_images/temporal_sequence/image_03.tif'", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence/image_01.tif"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence/image_02.tif"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence/image_03.tif"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 151, "task_desc": "Can you analyze the sequence of satellite images to assess the damage levels caused by a recent natural disaster? Please provide a detailed classification of the damage levels and a comprehensive interpretation of the affected areas. The image is stored in the directory: '/local/data/satellite/images/image_sequence_01.png'. These image is temporal sequence with 3 channels (Red, Green, Blue). User queries '/local/data/user_queries/query_01.txt'", "structured_plan": [{"agent": "Temporal_Image_Sequence_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/satellite/images/image_sequence_01.png"}, "outputs": ["classification_results_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["classification_results_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/user_queries/query_01.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 152, "task_desc": "In the aftermath of a natural disaster, assess the extent of damage by generating high-resolution images of the affected area after events through high-resolution and low-resolution images before event and low-resolution image after the event. Utilize these images to create a detailed change map that highlights significant alterations in the landscape, such as destroyed buildings or infrastructure changes. This information will aid in effective disaster response and recovery planning. Please provide the paths to the low-resolution image before the disaster, the high-resolution image before the disaster, and the low-resolution image after the disaster. The paths to the required images are as follows: 'data/images/low_res_before_disaster.jpg', 'data/images/high_res_before_disaster.jpg', 'data/images/low_res_after_disaster.jpg'. The metadata associated with these images is located at 'data/metadata/satellite_metadata.json'.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/images/low_res_before_disaster.jpg", "high_res_image_T_path": "data/images/high_res_before_disaster.jpg", "low_res_image_T_prime_path": "data/images/low_res_after_disaster.jpg", "metadata_path": "data/metadata/satellite_metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_1_path": "data/images/high_res_before_disaster.jpg", "image_2_path": "-0-"}, "outputs": ["change_map_path"]}], "source": "benchmark"} {"task_id": 153, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or earthquake, generate high-resolution satellite images of the affected area to assess damage and identify critical infrastructure. Utilize available low-resolution multi-spectral satellite images and relevant metadata to reconstruct detailed high-resolution images. Subsequently, perform geospatial object segmentation on these high-resolution images to identify and classify damaged structures, vehicles, and other key objects. This information will aid in coordinating emergency response efforts and resource allocation. Please provide the path to the low-resolution multi-spectral image data. The low-resolution multi-spectral image data is stored at '/data/satellite_images/low_res/multi_spectral_image_01.tif'. The metadata providing contextual details for the satellite image is located at '/data/satellite_images/metadata/contextual_metadata_01.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res/multi_spectral_image_01.tif", "metadata_path": "/data/satellite_images/metadata/contextual_metadata_01.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 154, "task_desc": "I need to assess the impact of a recent disaster on urban infrastructure. I have high-resolution satellite images taken before and after the event. Can you help identify any anomalies and then changes in the urban landscape, such as new constructions or damages to existing structures? Please use the images located at '/local/data/pre_disaster_image/high_res_satellite_pre_disaster.jpg' and '/local/data/post_disaster_image/high_res_satellite_post_disaster.jpg'. The pre-disaster image provides a baseline of the urban area before the event, while the post-disaster image captures the same area after the event, allowing for comparison and analysis of changes.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/pre_disaster_image/high_res_satellite_post_disaster.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/local/data/pre_disaster_image/high_res_satellite_pre_disaster.jpg", "image_2_path": "/local/data/post_disaster_image/high_res_satellite_post_disaster.jpg"}, "outputs": ["change_map_path"]}], "source": "benchmark"} {"task_id": 155, "task_desc": "In the aftermath of a natural disaster in an urban area, analyze satellite imagery to identify and describe anomalies that may indicate damage or changes in the environment. Use the anomaly detection process to generate a map highlighting potential areas of concern, and then provide a detailed contextual description of these anomalies to assist in disaster response and management efforts. Please ensure to input the path of the satellite image data for analysis. The satellite image data can be found at '/data/satellite_images/urban_area_post_disaster.tif'. The user query describing the task can be found at '/data/user_queries/disaster_response_query.txt'.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_post_disaster.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/disaster_response_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 156, "task_desc": "In the aftermath of a natural disaster, assess the extent of damage to buildings and identify anomalies in forested areas using high-resolution satellite imagery. Utilize pre- and post-disaster images to classify the level of damage to structures and generate a detailed damage classification map. Additionally, analyze the forest regions to detect any unusual patterns or changes, such as the presence of man-made structures or signs of forest health issues. Ensure that the analysis is comprehensive and provides actionable insights for disaster response and recovery efforts. The data paths for the required images are as follows: Pre-disaster image: '/data/satellite_images/pre_disaster_image.tif', Post-disaster image: '/data/satellite_images/post_disaster_image.tif', Forest area image: '/data/satellite_images/forest_area_image.tif'. These images are high-resolution satellite images used for damage assessment and anomaly detection.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 157, "task_desc": "Analyze the provided high-resolution satellite image to identify and highlight areas that are prone to landslides or have been affected by landslides. Additionally, assess the image for any anomalies in urban environments, particularly those related to disaster events. This will help in understanding and managing potential risks in both mountainous and urban areas. Please provide the path to the satellite image data for processing. The satellite image data can be found at the following paths: '/data/satellite_images/mountainous_terrain/landsat_image_2023.tif' for the mountainous terrain and '/data/satellite_images/urban_area/urban_image_2023.tif' for the urban area. These images are high-resolution TIFF files capturing the respective terrains.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain/landsat_image_2023.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area/urban_image_2023.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 158, "task_desc": "During a flood event, assess the impact on traffic by predicting water depths over a region and determining how these depths affect traffic speeds on road networks. Use the topography and precipitation data available at '/data/topography/topography_data.tif' and '/data/precipitation/precipitation_data.csv' to simulate the flood depths. The road network data is available at '/data/road_network/road_network_data.shp'.", "structured_plan": [{"agent": "Flood_depth_prediction", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"region_topography_path": "/data/topography/topography_data.tif", "precipitation_data_path": "/data/precipitation/precipitation_data.csv"}, "outputs": ["predicted_water_depths_path"]}, {"agent": "depth_speed_model", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_water_depths_path"]}, "inputs": {"road_network_path": "/data/road_network/road_network_data.shp", "flood_depth_path": "-0-"}, "outputs": ["traffic_speed_path"]}], "source": "benchmark"} {"task_id": 159, "task_desc": "Analyze satellite imagery to identify and map anomalies in forested areas that could indicate potential disaster events, such as wildfires or disease outbreaks. Use high-resolution images to detect deviations from normal forest patterns and generate an anomaly map highlighting areas of concern. Additionally, segment geospatial objects within the imagery to provide a comprehensive understanding of the affected regions. Please ensure the satellite image data is accessible at the specified paths: /local/data/satellite/image/forest_area_image.tif for the main input image representing a forest area, and /local/data/satellite/image/geospatial_area_image.tif for the high spatial resolution (HSR) satellite image containing the geospatial area of interest.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite/image/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite/image/geospatial_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 160, "task_desc": "Analyze the provided high-resolution satellite image to identify and map landslide-prone areas and segment various geospatial objects such as ships, airplanes, and vehicles. This analysis will help in assessing potential risks and planning disaster management strategies effectively. Please ensure the image is related to a landslide scenario and provide the path to the image data. Image path: '/data/satellite_images/landslide_scenario_image.tif'", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/landslide_scenario_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 161, "task_desc": "In the aftermath of a natural disaster, such as a hurricane or flood, assess the impact on infrastructure and identify areas requiring immediate attention. Utilize high-resolution satellite imagery to detect and highlight objects and structures affected by low-light conditions, ensuring accurate identification of damaged areas. Additionally, restore images degraded by adverse weather conditions to enhance clarity and detail, facilitating better decision-making for emergency response teams. Provide the path to the satellite imagery data for processing. The satellite imagery data can be found at the following paths: '/data/satellite_images/low_light_image_01.tif' for low-light conditions and '/data/satellite_images/weather_degraded_image_01.tif' for weather-degraded conditions. These images are high-resolution and captured in the aftermath of a natural disaster, providing crucial information for analysis.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_light_image_01.tif"}, "outputs": ["detections_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/weather_degraded_image_01.tif"}, "outputs": ["restored_image_path"]}], "source": "benchmark"} {"task_id": 162, "task_desc": "Generate a high-resolution satellite image of a region affected by a natural disaster using metadata and textual prompts. Then, analyze the generated image to estimate the number of people in the area, taking into account adverse weather conditions that may affect visibility. The descriptive caption and metadata are stored in local files, and the generated image will be analyzed for crowd estimation. The input files are stored at '/local/data/captions/disaster_region_caption.txt' and '/local/data/metadata/disaster_region_metadata.json' for captions and metadata respectively.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/local/data/captions/disaster_region_caption.txt", "metadata_path": "/local/data/metadata/disaster_region_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 163, "task_desc": "Analyze the provided low-resolution multi-spectral satellite image to identify potential disaster-related anomalies in both urban and forest environments. Use the image data to generate a high-resolution image, then detect any unusual patterns or objects that may indicate disaster events such as wildfires in forests or structural damage in urban areas. Ensure to segment and classify geospatial objects to assist in disaster response and management. The low-resolution image data is located at '/data/satellite_images/low_res_image.tif'. The metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata_vector.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image.tif", "metadata_path": "/data/satellite_images/metadata_vector.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 3, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 164, "task_desc": "Analyze high-resolution satellite images to identify and classify anomalies in both forest and urban environments that may indicate potential disaster events. Use the images to detect unusual patterns or objects, such as signs of wildfires in forests or structural damages in urban areas, and classify the detected features into relevant categories. Provide a detailed segmentation map highlighting geospatial objects and anomalies, along with confidence scores for each identified feature. Ensure the analysis covers both natural and man-made anomalies to support effective disaster management and response planning. The data paths for the required images and additional files are as follows: - Forest area image: /data/satellite_images/forest_area_image.tif - Urban area image: /data/satellite_images/urban_area_image.tif - High spatial resolution (HSR) satellite image: /data/satellite_images/hsr_geospatial_image.tif - Main satellite image with spectral bands: /data/satellite_images/spectral_bands_image.tif - Binary mask array: /data/masks/binary_mask_array.npy", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/hsr_geospatial_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/spectral_bands_image.tif", "mask_path": "/data/masks/binary_mask_array.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 165, "task_desc": "In the aftermath of a forest disaster, such as a forest fire, analyze high-resolution satellite images to identify and map anomalies within the forest environment. This includes detecting man-made structures, water bodies, vehicles, and signs of forest health issues. Additionally, assess the impact of the disaster by generating a high-resolution image of the affected area using available low and high-resolution images taken before the disaster. Use this generated image to further classify and categorize visible features and assess potential landslide risks in mountainous regions. Ensure all necessary data is available locally for processing. The data paths are as follows: - Low-resolution satellite image at time T: '/data/satellite_images/low_res_image_T.jpg'- High-resolution satellite image at time T: '/data/satellite_images/high_res_image_T.jpg'- Low-resolution satellite image at time T': '/data/satellite_images/low_res_image_T_prime.jpg'- Metadata associated with the satellite images: '/data/metadata/satellite_metadata.json'.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/data/satellite_images/low_res_image_T.jpg", "high_res_image_T_path": "/data/satellite_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "/data/satellite_images/low_res_image_T_prime.jpg", "metadata_path": "/data/metadata/satellite_metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Landslide_Segmentation", "step": 3, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 166, "task_desc": "In the aftermath of a natural disaster, analyze high-resolution satellite images to identify and describe the impact on infrastructure and environment. Use the images to detect objects and assess the extent of damage, even these images are in challenging conditions like low-light or fog. Provide a detailed contextual description of the affected areas to aid in disaster management and recovery efforts. The images are located at the specified paths: '/local/data/satellite/images/foggy_image_01.jpg', '/local/data/satellite/images/low_light_image_01.jpg'. These images are high-resolution satellite captures, with 'foggy_image_01.jpg' taken in foggy conditions and 'low_light_image_01.jpg' taken in low-light conditions. User queries are stored at '/local/data/queries/user_query.txt'", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite/images/foggy_image_01.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite/images/low_light_image_01.jpg"}, "outputs": ["detections_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["detected_objects_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/local/data/queries/user_query.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["detections_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/local/data/queries/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 167, "task_desc": "Analyze high-resolution satellite images to identify and classify potential disaster-related anomalies in various environments. Use the image data to detect and highlight unusual patterns or features that may indicate landslides in mountainous regions, urban anomalies related to disaster events, or forest anomalies such as wildfires. Additionally, classify the detected features into specific categories to aid in disaster management and response planning. Please provide the path to the high-resolution image data for processing. The data paths are as follows: - Mountainous terrain image: '/data/satellite_images/mountainous_terrain/high_res_image_01.tif'- Urban area image: '/data/satellite_images/urban_area/high_res_image_02.tif'- Forest area image: '/data/satellite_images/forest_area/high_res_image_03.tif'- Sentinel-2 spectral bands image: '/data/satellite_images/sentinel_2/spectral_bands_image_04.tif'- Binary mask for Sentinel-2 image: '/data/satellite_images/sentinel_2/mask_image_04.tif'.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain/high_res_image_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area/high_res_image_02.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area/high_res_image_03.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Multi_Spectral_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/sentinel_2/spectral_bands_image_04.tif", "mask_path": "/data/satellite_images/sentinel_2/mask_image_04.tif"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 168, "task_desc": "There are some high-resolution satellite images, you need to finish the following tasks. Identify and segment various geospatial objects such as buildings, vehicles, and other structures. Additionally, assess the image for any anomalies that may indicate potential disaster events, such as landslides in mountainous regions, urban anomalies in city areas, or forest anomalies like wildfires. Provide a detailed segmentation map and anomaly detection results to aid in disaster management and response planning. You should give suitable image to the corresponding suitable model based on image description and model description. One model will only take one image as input in this task. The data paths for the images and preprocessing requirements are as follows: - High spatial resolution (HSR) satellite image for geospatial object segmentation: '/data/satellite_images/geospatial_area_of_interest.jpg'- High-resolution satellite image of an urban area for urban anomaly detection: '/data/satellite_images/urban_area.jpg'- High-resolution aerial or satellite image capturing mountainous terrain for landslide segmentation: '/data/satellite_images/mountainous_terrain.jpg'- Main input image representing a forest area for anomaly detection in forests: '/data/satellite_images/forest_area.jpg'.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_of_interest.jpg"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 169, "task_desc": "Analyze the impact of a recent disaster on a forested area by identifying anomalies and identify changes in the landscape. Additionally, generate a detailed segmentation map to classify different geospatial objects within the affected area, and provide a comprehensive description of the scene to assist in disaster management efforts. The data paths for the required images and files are as follows: 'data/satellite_images/pre_disaster_image.tif' for the satellite image before the disaster, 'data/satellite_images/post_disaster_image.tif' for the satellite image after the disaster, and 'data/user_queries/disaster_management_query.txt' for the user query description.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/pre_disaster_image.tif", "image_2_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [2], "dependence_content": {"2": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-2-", "user_query_path": "data/user_queries/disaster_management_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 170, "task_desc": "Analyze the provided high-resolution satellite imagery to identify and describe any forest anomalies or unusual patterns that could indicate potential disaster risks. Provide a detailed contextual description of the identified anomalies to assist in disaster management and mitigation efforts. The data paths for the required images and files are as follows: - Forest area image: /data/satellite_images/forest_area_image.tif - High spatial resolution (HSR) satellite image: /data/satellite_images/hsr_satellite_image.tif - User queries /data/user_queries.json", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/hsr_satellite_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries.json"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 171, "task_desc": "In the aftermath of a natural disaster, this task involves assessing the extent of damage to urban infrastructure and surrounding forest areas. Below are details. Utilize high-resolution satellite imagery captured before and after the event to identify changes in the landscape, such as newly constructed buildings or alterations in existing structures. Additionally, detect and map anomalies in forest regions, highlighting areas affected by the disaster, such as forest fires or other disturbances. Provide a comprehensive analysis of the affected regions, including a detailed segmentation of geospatial objects and a classification of building damage severity. Ensure to include the path to the pre- and post-disaster images for processing. The data paths are as follows: - Pre-disaster image: '/data/satellite_images/pre_disaster_image_2023.tif' - Post-disaster image: '/data/satellite_images/post_disaster_image_2023.tif' - Forest area image: '/data/satellite_images/forest_area_image_2023.tif' - High spatial resolution image: '/data/satellite_images/hsr_image_2023.tif'", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/data/satellite_images/pre_disaster_image_2023.tif", "image_2_path": "/data/satellite_images/post_disaster_image_2023.tif"}, "outputs": ["change_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image_2023.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/hsr_image_2023.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Building_damage_assessment", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image_2023.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image_2023.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 172, "task_desc": "Analyze satellite imagery to identify and describe anomalies in forest and urban environments that could indicate potential disaster risks, such as landslides, diseased trees, or unexpected urban developments. Provide a detailed contextual description of these anomalies to aid in disaster management and response planning. The data paths for the required images and user queries are as follows: - Forest area satellite image: /data/satellite_images/forest_area_image.tif - Urban area satellite image: /data/satellite_images/urban_area_image.tif - User query for forest analysis: /data/user_queries/forest_query.txt - User query for urban analysis: /data/user_queries/urban_query.txt", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/user_queries/forest_query.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/urban_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 173, "task_desc": "In the aftermath of a natural disaster, your task is to assess the impact on both urban and forest environments by analyzing high-resolution satellite images taken before and after the event. Identify and map any significant changes in infrastructure, such as new constructions or damage, and detect anomalies in urban areas that deviate from typical patterns. Additionally, highlight any irregularities in forest regions, such as signs of forest fires or other disruptions. Utilize the temporal sequence of images to classify the affected areas. Please provide the path to the high-resolution images captured before and after the disaster for analysis. The paths to the required data are as follows: - High-resolution satellite image before the disaster: '/data/satellite_images/before_disaster_image_01.tif'- High-resolution satellite image after the disaster: '/data/satellite_images/after_disaster_image_02.tif'- High-resolution satellite image of an urban area: '/data/satellite_images/urban_area_image_03.tif'- High-resolution satellite image of a forest area: '/data/satellite_images/forest_area_image_04.tif'- Temporal sequence of satellite images: '/data/satellite_images/temporal_sequence_images_05.tif'.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/data/satellite_images/before_disaster_image_01.tif", "image_2_path": "/data/satellite_images/after_disaster_image_02.tif"}, "outputs": ["change_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image_03.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image_04.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/data/satellite_images/temporal_sequence_images_05.tif"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 174, "task_desc": "In the aftermath of a natural disaster, it is crucial to assess the extent of damage and identify areas that require immediate attention. Using satellite imagery, generate high-resolution images from low-resolution multi-spectral data to analyze potential landslide-prone areas in mountainous regions. Additionally, perform geospatial object segmentation to identify and classify objects or scenes within the imagery. For urban areas affected by the disaster, detect anomalies that deviate from typical urban patterns to localize and analyze irregular features. Ensure that the necessary data is available for processing and analysis. The data paths are as follows: low-resolution satellite image input is located at '/data/satellite_images/low_res_image.tif', metadata vector providing contextual details for the satellite image is located at '/data/satellite_images/metadata.json', high spatial resolution (HSR) satellite image containing the geospatial area of interest is located at '/data/satellite_images/hsr_image.tif', and high-resolution satellite image of an urban area is located at '/data/satellite_images/high_res_urban_image.tif'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image.tif", "metadata_path": "/data/satellite_images/metadata.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/hsr_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/high_res_urban_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 175, "task_desc": "Analyze satellite imagery to identify and highlight potential disaster-prone areas in various environments. Use low-resolution multi-spectral images to generate high-resolution images, which can then be used to detect anomalies in urban, forest, and mountainous regions. The goal is to identify areas at risk of disasters such as landslides, urban anomalies, and forest anomalies. Provide detailed anomaly maps for each environment to assist in disaster management and intervention planning. Ensure the data paths for the required images are correctly set for processing. The data paths are as follows: - Low-resolution satellite image input: '/data/satellite_images/low_res_image_01.tif'- Metadata vector: '/data/satellite_images/metadata_01.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image_01.tif", "metadata_path": "/data/satellite_images/metadata_01.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 3, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 176, "task_desc": "In the aftermath of a severe storm, assess the impact on a coastal city by analyzing satellite imagery. Use high-resolution satellite images to detect objects and count crowds in areas affected by fog and low-light conditions. This will help in understanding the extent of damage, identifying areas with high population density, and planning effective relief operations. The satellite images are stored locally and can be accessed at the following paths: 'data/satellite_images/foggy_conditions_image.jpg' for images captured in foggy conditions, 'data/satellite_images/low_light_conditions_image.jpg' for images captured under low-light conditions, and 'data/satellite_images/crowd_scene_image.jpg' for images containing crowd scenes.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/foggy_conditions_image.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/low_light_conditions_image.jpg"}, "outputs": ["detections_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/crowd_scene_image.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 177, "task_desc": "In the aftermath of a natural disaster, assess the extent of damage to urban infrastructure and identify any anomalies in the affected areas. Utilize high-resolution satellite imagery taken before and after the disaster to detect changes in the landscape, such as new constructions or alterations in existing structures. Additionally, perform a detailed segmentation of geospatial objects to understand the impact on specific elements like buildings and vehicles. Provide a comprehensive analysis of the situation, highlighting areas of significant change and potential hazards, to aid in effective disaster response and recovery efforts. The data paths for the required images and preprocessing requirements are as follows: 'data/satellite_images/pre_disaster_image.tif' for the pre-disaster image, 'data/satellite_images/post_disaster_image.tif' for the post-disaster image, 'data/remote_sensing_images/image_1.tif' for the first remote sensing image, 'data/remote_sensing_images/image_2.tif' for the second remote sensing image, 'data/high_res_images/urban_area_image.tif' for the high-resolution satellite image of the urban area", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.tif", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/remote_sensing_images/image_1.tif", "image_2_path": "data/remote_sensing_images/image_2.tif"}, "outputs": ["change_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/high_res_images/urban_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 178, "task_desc": "In the aftermath of a natural disaster, assess the damage to buildings and identify any anomalies in the affected forest and mountainous regions. Utilize pre- and post-disaster satellite imagery to generate high-resolution images of the impacted areas. Analyze these images to classify the extent of building damage, detect anomalies in forest environments, and identify landslide-prone or affected areas. Provide detailed maps indicating the likelihood of anomalies and the classification of building damage to aid in disaster response and recovery efforts. The data paths for the required images and metadata are as follows: - Low-resolution satellite image at time T: '/data/satellite_images/low_res_image_T.jpg'- High-resolution satellite image at time T: '/data/satellite_images/high_res_image_T.jpg'- Low-resolution satellite image at time T': '/data/satellite_images/low_res_image_T_prime.jpg'- Metadata associated with the satellite images: '/data/satellite_images/metadata.json'- Pre-disaster satellite image: '/data/satellite_images/pre_disaster_image.jpg'- Post-disaster satellite image: '/data/satellite_images/post_disaster_image.jpg'- Main input image for forest anomaly detection: '/data/satellite_images/forest_image.jpg'- High-resolution image for landslide segmentation: '/data/satellite_images/mountainous_terrain_image.jpg'.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/data/satellite_images/low_res_image_T.jpg", "high_res_image_T_path": "/data/satellite_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "/data/satellite_images/low_res_image_T_prime.jpg", "metadata_path": "/data/satellite_images/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_disaster_image.jpg", "post_disaster_image_path": "/data/satellite_images/post_disaster_image.jpg"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_image.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 179, "task_desc": "There is a low-resolution image. You need to help me reconstruct this low-resolution image for the downstream task. Then, I want you to analyze the generated high-resolution satellite imagery to identify and classify geospatial objects and anomalies in urban areas affected by recent disaster events. Provide a detailed segmentation map highlighting different object types and an anomaly map indicating unusual patterns or features. Additionally, generate a comprehensive contextual description of the scene to assist in disaster response and management efforts. Please ensure the satellite image data is accessible from the specified path. The low-resolution satellite image is located at '/data/satellite_images/low_res_image.jpg', and the metadata vector is available at '/data/satellite_images/metadata_vector.json'. The user query for the GeoChat model is located at '/data/satellite_images/user_query.txt'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/low_res_image.jpg", "metadata_path": "/data/satellite_images/metadata_vector.json"}, "outputs": ["high_resolution_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [0], "dependence_content": {"0": ["high_resolution_image_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "/data/satellite_images/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 180, "task_desc": "Assess the impact of a recent disaster on urban infrastructure by generating high-resolution images of the affected area before and after the event. Additionally, identify any anomalies in the urban environment that may have resulted from the disaster. Ensure to provide the paths to the pre-disaster low and high-resolution images, as well as the post-disaster low-resolution image, to facilitate the analysis. The data paths are as follows: 'data/pre_disaster_low_res_image_T.jpg', 'data/pre_disaster_high_res_image_T.jpg', 'data/post_disaster_low_res_image_T_prime.jpg', 'data/metadata.json'.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/pre_disaster_low_res_image_T.jpg", "high_res_image_T_path": "data/pre_disaster_high_res_image_T.jpg", "low_res_image_T_prime_path": "data/post_disaster_low_res_image_T_prime.jpg", "metadata_path": "data/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 181, "task_desc": "Analyze high-resolution satellite images taken before and after a natural disaster to identify changes in the landscape. You need to detect anomalies in forested areas, such as signs of forest fires or other disruptions, and identifying potential landslide-prone regions. Additionally, assess urban areas for unexpected changes or damages caused by the disaster. Provide a detailed report highlighting the affected regions and categorizing the types of changes observed. The data paths for the images are as follows: 'data/satellite_images/pre_disaster_image.tif' for the image taken before the disaster, and 'data/satellite_images/post_disaster_image.tif' for the image taken after the disaster. These images are high-resolution satellite images capturing the same geographical area at different times.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/pre_disaster_image.tif", "image_2_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 3, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 182, "task_desc": "Analyze the high-resolution satellite imagery to identify and describe any anomalies or unusual patterns in urban areas that could indicate potential disaster events. Use the image data to detect and segment geospatial objects, and generate a detailed contextual description of the findings, highlighting any significant deviations from typical urban patterns that may require further investigation or intervention. The high-resolution satellite image data is stored at '/data/satellite_images/urban_area_image.tif', user query is located at '/data/user_queries/user_query.txt'.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [1], "dependence_content": {"1": ["segmentation_map_image_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/user_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 183, "task_desc": "Analyze satellite imagery to identify and describe potential landslide-prone areas and urban anomalies. Use the available data to generate detailed contextual descriptions of these regions, highlighting any deviations from typical patterns that may indicate a risk of disaster. The data paths for the required images and user queries are as follows: - High-resolution aerial or satellite image capturing mountainous terrain: './data/satellite_images/mountainous_terrain_image_01.tif'- High-resolution satellite image of an urban area: './data/satellite_images/urban_area_image_01.tif'- User query for landslide-prone area analysis: './data/user_queries/landslide_query.txt'- User query for urban anomaly detection: './data/user_queries/urban_anomaly_query.txt'", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "./data/satellite_images/mountainous_terrain_image_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "./data/satellite_images/urban_area_image_01.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 2, "dependence": [0], "dependence_content": {"0": ["anomaly_map_path"]}, "inputs": {"image_path": "-0-", "user_query_path": "./data/user_queries/landslide_query.txt"}, "outputs": ["contextual_description_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "./data/user_queries/urban_anomaly_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 184, "task_desc": "In the event of a natural disaster, such as a severe storm or hurricane, generate high-resolution satellite images of the affected area using available metadata and textual prompts. Utilize these images to detect objects and assess the extent of damage, including identifying infrastructure and vehicles. Additionally, estimate the crowd density in public areas to aid in emergency response and resource allocation. Ensure that the images are restored to enhance clarity and detail, especially in adverse weather conditions, to improve the accuracy of the analysis. The data paths for the task are as follows: 'data/captions/disaster_caption.txt' for the descriptive caption, 'data/metadata/disaster_metadata.json' for the metadata, 'data/images/weather_degraded_image.jpg' for the weather-degraded image, and 'data/images/restored_image.jpg' for the restored image.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "data/captions/disaster_caption.txt", "metadata_path": "data/metadata/disaster_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 3, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 185, "task_desc": "Analyze high-resolution satellite images taken before and after a natural disaster to identify and map changes in urban and forest environments. Detect anomalies such as new constructions, damaged infrastructure, or unusual patterns in the landscape. Provide a detailed report on the affected areas, highlighting significant changes and potential risks to aid in disaster response and recovery efforts. Using the above analysis, provide a contextual description of the urban scenario. The data paths for the images are as follows: 'data/satellite_images/pre_disaster_image.tif' for the image taken before the disaster, 'data/satellite_images/post_disaster_image.tif' for the image taken after the disaster, 'data/satellite_images/urban_area_image.tif' for the urban area image, 'data/satellite_images/forest_area_image.tif' for the forest area image, and 'data/user_queries/disaster_response_query.txt' for the user query describing the task.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/pre_disaster_image.tif", "image_2_path": "data/satellite_images/post_disaster_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "data/user_queries/disaster_response_query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 186, "task_desc": "Analyze the impact of a recent disaster in an urban area by generating high-resolution images of the location before and after the event. Use these images to identify and highlight anomalies in the urban environment, such as damaged infrastructure or unexpected changes. Additionally, segment and classify geospatial objects within the affected area to assess the extent of the damage. Provide a detailed contextual description of the anomalies and changes observed in the urban landscape. The data paths for the required inputs are as follows: 'low_res_image_T_path': '/data/satellite_images/low_res_image_T.jpg', 'high_res_image_T_path': '/data/satellite_images/high_res_image_T.jpg', 'low_res_image_T_prime_path': '/data/satellite_images/low_res_image_T_prime.jpg', 'metadata_path': '/data/satellite_images/metadata.json', 'image_path': '/data/satellite_images/high_res_image.jpg', 'user_query_path': '/data/user_queries/query.txt'.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/data/satellite_images/low_res_image_T.jpg", "high_res_image_T_path": "/data/satellite_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "/data/satellite_images/low_res_image_T_prime.jpg", "metadata_path": "/data/satellite_images/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["high_res_image_T_prime_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [1], "dependence_content": {"1": ["anomaly_map_path"]}, "inputs": {"image_path": "-1-", "user_query_path": "/data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 187, "task_desc": "In the aftermath of a natural disaster, assess the impact on urban infrastructure by analyzing high-resolution satellite images taken before and after the event. Generate a detailed change map to identify areas with significant alterations, such as damaged buildings or roads. Additionally, perform geospatial object segmentation to classify and map various objects within the affected area, and detect any anomalies that deviate from typical urban patterns. Provide a comprehensive contextual description of the scene to aid in disaster response and recovery efforts. The data paths for the images and other necessary files are as follows: 'data/satellite_images/before_event_image.tif' for the image taken before the event, 'data/satellite_images/after_event_image.tif' for the image taken after the event, and 'data/user_queries/query.txt' for the user query description.", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/satellite_images/before_event_image.tif", "image_2_path": "data/satellite_images/after_event_image.tif"}, "outputs": ["change_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["change_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "GeoChat", "step": 3, "dependence": [2], "dependence_content": {"2": ["anomaly_map_path"]}, "inputs": {"image_path": "-2-", "user_query_path": "data/user_queries/query.txt"}, "outputs": ["contextual_description_path"]}], "source": "benchmark"} {"task_id": 188, "task_desc": "Assume that there is a severe storm happened in a coastal city, can you help me synthesize the high resolution satellite image in the coastal city after the storm with a lot of people and in foggy environment in the night. Utilize satellite imagery to identify and count the number of people in crowded areas. Additionally, even under challenging weather conditions like fog or low light, detect any objects that might pose a risk to public safety, such as debris or damaged infrastructure. This information will be crucial for coordinating emergency response efforts and ensuring the safety of the affected population. Please provide the path to the satellite image data for analysis. The data paths are as follows: - Caption data: /local/data/captions/storm_caption.txt - Metadata: /local/data/metadata/storm_metadata.json", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/local/data/captions/storm_caption.txt", "metadata_path": "/local/data/metadata/storm_metadata.json"}, "outputs": ["generated_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}, {"agent": "Low-Light_Object_Detection", "step": 3, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 189, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a densely populated urban area. The image is in low quality as it is affected by the weather. Utilize high-resolution satellite imagery to identify and count the number of people in crowded areas affected by the storm. Additionally, detect and describe any significant objects or structures that may have been damaged or are obstructing roads due to the storm's impact. Please ensure the imagery is clear and free from weather-related distortions to improve the accuracy of the analysis. The satellite image is located at the path: /path/to/impacted_image.jpg", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/path/to/impacted_image.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 190, "task_desc": "Assess the damage caused by a recent disaster by analyzing pre- and post-disaster high-resolution satellite images. Generate a detailed map that classifies the extent of damage to buildings and infrastructure. Next, identify any significant changes in the landscape using the remote sensing images. Additionally, segment and categorize various geospatial objects within the affected area to understand the impact on different elements such as vehicles, ships, and airplanes. Ensure to include any potential landslide-prone areas in the analysis for a comprehensive disaster management report. The data paths for the required images are as follows: 'data/pre_disaster_image.tif' for the pre-disaster satellite image, 'data/post_disaster_image.tif' for the post-disaster satellite image, 'data/image_1.tif' for the first remote sensing image, 'data/image_2.tif' for the second remote sensing image, 'data/geospatial_image.tif' for the high spatial resolution satellite image, and 'data/landslide_image.tif' for the high-resolution image capturing mountainous terrain.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "data/pre_disaster_image.tif", "post_disaster_image_path": "data/post_disaster_image.tif"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Change_Mapping_and_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "data/image_1.tif", "image_2_path": "data/image_2.tif"}, "outputs": ["change_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/geospatial_image.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Landslide_Segmentation", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "data/landslide_image.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 191, "task_desc": "Assess the impact of a recent disaster on buildings and identify potential landslide-prone areas using satellite imagery. Utilize high-resolution images captured before and after the disaster to classify the extent of building damage and detect anomalies in mountainous regions that may indicate landslides. Ensure to incorporate temporal sequences of images to enhance the accuracy of the classification and segmentation tasks. Please provide the path to the pre-disaster and post-disaster images, as well as any relevant metadata, to facilitate the analysis. The data paths are as follows: 'data/pre_disaster_image.jpg' for the pre-disaster image, 'data/post_disaster_image.jpg' for the post-disaster image, 'data/low_res_image_T.jpg' for the low-resolution image at time T, 'data/high_res_image_T.jpg' for the high-resolution image at time T, 'data/low_res_image_T_prime.jpg' for the low-resolution image at time T', 'data/metadata.json' for the metadata, and 'data/temporal_sequence_images/image_sequence.png' for the sequence of images.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "data/low_res_image_T.jpg", "high_res_image_T_path": "data/high_res_image_T.jpg", "low_res_image_T_prime_path": "data/low_res_image_T_prime.jpg", "metadata_path": "data/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}, {"agent": "Building_damage_assessment", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "data/pre_disaster_image.jpg", "post_disaster_image_path": "data/post_disaster_image.jpg"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [0], "dependence_content": {"0": ["damage_classification_map_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["anomaly_map_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "data/temporal_sequence_images/image_sequence.png"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 192, "task_desc": "In the aftermath of a severe storm, assess the impact on urban infrastructure by analyzing satellite images. Enhance the clarity of weather-degraded images to identify and count the number of people in affected areas, and detect objects such as vehicles and buildings in low-light conditions. Use the path to the satellite images to begin the analysis. The satellite images are stored at '/data/satellite_images/storm_aftermath_image1.jpg', '/data/satellite_images/storm_aftermath_image2.jpg'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/storm_aftermath_image1.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/satellite_images/storm_aftermath_image2.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 3, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 4, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}, {"agent": "Low-Light_Object_Detection", "step": 5, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 193, "task_desc": "In the aftermath of a severe storm, we need to assess the impact on a coastal city. Generate high-resolution satellite images of the affected area using available metadata and textual prompts. Once the images are generated, enhance them to restore clarity and detail lost due to adverse weather conditions like rain and haze. Detect and count the number of people in crowded areas to assist in emergency response and resource allocation. Additionally, identify any objects or obstacles in low-light conditions to ensure safe navigation for rescue operations. Please provide the path to the metadata and textual prompts for image generation. The metadata and textual prompts are stored in the following paths: '/data/metadata/metadata_vector.json' and '/data/text_prompts/descriptive_caption.txt'.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/data/text_prompts/descriptive_caption.txt", "metadata_path": "/data/metadata/metadata_vector.json"}, "outputs": ["generated_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [0], "dependence_content": {"0": ["generated_image_path"]}, "inputs": {"weather_degraded_image_path": "-0-"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 3, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 194, "task_desc": "In the aftermath of a severe storm, assess the impact on a coastal city by analyzing satellite images. Use the available high-resolution satellite imagery to identify and count the number of people in crowded areas affected by the storm, even under adverse weather conditions like rain or fog. Additionally, detect any objects or debris that may pose a hazard to the public. Ensure that the images are restored to enhance clarity and detail before performing these analyses. Provide the path to the satellite images for processing. The satellite images are stored at '/local/data/satellite_images/storm_aftermath_image1.jpg', '/local/data/satellite_images/storm_aftermath_image2.jpg', and '/local/data/satellite_images/storm_aftermath_image3.jpg'.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/satellite_images/storm_aftermath_image1.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/satellite_images/storm_aftermath_image2.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/satellite_images/storm_aftermath_image3.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 3, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 4, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 5, "dependence": [2], "dependence_content": {"2": ["restored_image_path"]}, "inputs": {"image_path": "-2-"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Low-Light_Object_Detection", "step": 6, "dependence": [0], "dependence_content": {"0": ["restored_image_path"]}, "inputs": {"image_path": "-0-"}, "outputs": ["detections_path"]}, {"agent": "Low-Light_Object_Detection", "step": 7, "dependence": [1], "dependence_content": {"1": ["restored_image_path"]}, "inputs": {"image_path": "-1-"}, "outputs": ["detections_path"]}, {"agent": "Low-Light_Object_Detection", "step": 8, "dependence": [2], "dependence_content": {"2": ["restored_image_path"]}, "inputs": {"image_path": "-2-"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 195, "task_desc": "In the aftermath of a severe weather event, such as a hurricane or snowstorm, assess the impact on urban areas by analyzing satellite imagery and on-ground photos. Use the available data to restore clarity to weather-degraded images, detect objects and infrastructure in low-light and foggy conditions, and estimate crowd density in affected regions. This information will aid in coordinating emergency response efforts and resource allocation for disaster relief operations. The data paths for the required inputs are as follows: 1. Weather-degraded images: '/data/weather_degraded_images/image1.jpg' 2. Low-light condition images: '/data/low_light_images/image2.jpg' 3. Foggy condition images: '/data/foggy_images/image3.jpg' 4. Crowd scene images: '/data/crowd_images/image4.jpg'. These images include satellite and on-ground photos affected by various weather conditions, which will be used for image restoration, object detection, and crowd counting.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/data/weather_degraded_images/image1.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Low-Light_Object_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/low_light_images/image2.jpg"}, "outputs": ["detections_path"]}, {"agent": "Foggy_Scenario_Object_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/foggy_images/image3.jpg"}, "outputs": ["detected_objects_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/crowd_images/image4.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 196, "task_desc": "Analyze the provided high-resolution satellite images to identify and segment potential landslide-prone areas, detect anomalies in forest environments such as wildfires or diseased trees, and highlight irregularities in urban settings that may indicate disaster events. Additionally, perform geospatial object segmentation to distinguish various objects within the imagery. Ensure to use the appropriate models for each scenario to obtain detailed anomaly maps and segmentation results. The data paths for the images are as follows: the image to segment potential landslide-prone areas is located at '/data/satellite_images/mountainous_terrain_image.tif', the image to detect anomalies in forest environments is at '/data/satellite_images/forest_area_image.tif', the image to highlight irregularities in urban settings is at '/data/satellite_images/urban_area_image.tif', and the image to perform geospatial object segmentation is at '/data/satellite_images/geospatial_area_image.tif'. ", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/mountainous_terrain_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/forest_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/urban_area_image.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 3, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/geospatial_area_image.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 197, "task_desc": "Can you generate a high-resolution satellite image of a specific location during a recent natural disaster event using the available metadata and textual information? The data needed for this task includes a descriptive caption and normalized numerical metadata. The caption provides context for the image to be generated, while the metadata offers additional contextual information to guide the image generation. The caption data can be found at '/local/data/satellite_image_generation/caption.txt' and the metadata is located at '/local/data/satellite_image_generation/metadata.json'.", "structured_plan": [{"agent": "Metadata_and_Text_Prompt_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"caption_path": "/local/data/satellite_image_generation/caption.txt", "metadata_path": "/local/data/satellite_image_generation/metadata.json"}, "outputs": ["generated_image_path"]}], "source": "benchmark"} {"task_id": 198, "task_desc": "Can you analyze the satellite image sequences to identify areas affected by recent natural disasters, like floods or wildfires, and classify them accordingly? The data is located at /local/data/satellite/images/sequence1.png. These images are temporal sequences, each with 3 channels (Red, Green, Blue).", "structured_plan": [{"agent": "Temporal_Image_Sequence_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/satellite/images/sequence1.png"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 199, "task_desc": "I have a set of low-resolution satellite images of a region affected by a recent hurricane, taken at different times. I also have a high-resolution image from before the hurricane. Can you help me generate a high-resolution image of the region after the hurricane using these images? The data is located at /local/path/to/hurricane_images/low_res_image_T.jpg for the low-resolution image at time T, /local/path/to/hurricane_images/high_res_image_before.jpg for the high-resolution image before the hurricane, /local/path/to/hurricane_images/low_res_image_T_prime.jpg for the low-resolution image at time T', and /local/path/to/hurricane_images/metadata.json for the metadata associated with the satellite images.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/local/path/to/hurricane_images/low_res_image_T.jpg", "high_res_image_T_path": "/local/path/to/hurricane_images/high_res_image_before.jpg", "low_res_image_T_prime_path": "/local/path/to/hurricane_images/low_res_image_T_prime.jpg", "metadata_path": "/local/path/to/hurricane_images/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}], "source": "benchmark"} {"task_id": 200, "task_desc": "Can you analyze the satellite images from the recent data collection and identify any potential landslide-prone areas? The images are stored in the following files: '/data/satellite_images/image1.tif'. These images are high-resolution satellite captures of mountainous terrain. I need a detailed anomaly map highlighting these areas for further assessment.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/satellite_images/image1.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 201, "task_desc": "Can you analyze the satellite images from the recent natural disaster and identify the affected areas by classifying the images into different categories? The images are located at /local/data/satellite/images/sentinel2_image.tif. The binary mask array specifying which regions of the satellite image are visible (value 0) or masked (value 1) during the inference process is located at /local/data/satellite/masks/visibility_mask.npy.", "structured_plan": [{"agent": "Multi_Spectral_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite/images/sentinel2_image.tif", "mask_path": "/local/data/satellite/masks/visibility_mask.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 202, "task_desc": "Can you identify and highlight any unusual patterns or anomalies in urban areas from the satellite image stored at '/data/urban_images/city1_image1.jpg'? I'm interested in understanding if there are any unexpected features or changes in these urban environments. The data consists of high-resolution satellite images of various urban areas, capturing different cities and their unique layouts.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/urban_images/city1_image1.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 203, "task_desc": "Can you predict the precipitation patterns for the next few hours using the radar data from 'data/radar_observations.csv'? I'm interested in understanding how the weather might change shortly to prepare for any potential flooding or severe weather conditions. The radar data is stored in 'data/radar_observations.csv', which contains time-stamped radar echo frames representing observed precipitation data over time.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "data/radar_observations.csv"}, "outputs": ["predicted_radar_echo_frames_path"]}], "source": "benchmark"} {"task_id": 204, "task_desc": "Can you identify and list the objects detected in images taken during foggy weather conditions? The images are stored in the folder located at '/local/data/foggy_images/image1.jpg'. These images are captured in foggy weather conditions and are used to test object detection models under such scenarios.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/foggy_images/image1.jpg"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 205, "task_desc": "Can you analyze the satellite images from before and after the recent hurricane to determine the extent of damage to buildings in the affected city? The images are located at /data/satellite_images/pre_hurricane/image1_pre.png and /data/satellite_images/post_hurricane/image1_post.png. The pre-hurricane image (image1_pre.png) captures the city before the hurricane, while the post-hurricane image (image1_post.png) shows the city after the hurricane, allowing for a comparative analysis to assess building damage.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_hurricane/image1_pre.png", "post_disaster_image_path": "/data/satellite_images/post_hurricane/image1_post.png"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 206, "task_desc": "I need to estimate the number of people in a crowded outdoor event captured in a series of images taken during different weather conditions like rain and snow. Can you help me analyze these images to get an accurate crowd count? The images are stored in the directory: /data/event_images/. The specific image files are: /data/event_images/image1_rain.jpg, /data/event_images/image2_snow.jpg, and /data/event_images/image3_clear.jpg. These images represent different weather conditions and will be used for analysis.", "structured_plan": [{"agent": "Crowd_Counting_in_Adverse_Weather", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/event_images/image1_rain.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/event_images/image2_snow.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}, {"agent": "Crowd_Counting_in_Adverse_Weather", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/event_images/image3_clear.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 207, "task_desc": "Can you analyze the forest imagery data located at '/local/data/forest_images/forest_image_01.jpg' and '/local/data/forest_images/forest_image_02.jpg' and provide an anomaly map that highlights any unusual patterns or structures, such as man-made objects, water bodies, or signs of forest health issues? The data consists of high-resolution images capturing various forest areas, which are essential for detecting anomalies effectively.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/forest_images/forest_image_01.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/forest_images/forest_image_02.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 208, "task_desc": "I have a set of images captured during a recent storm, and they are quite unclear due to the heavy rain and fog. Could you help me enhance these images to improve their clarity and detail? The images are located at /local_machine_path/storm_images/image1.jpg. These images are affected by weather conditions and need enhancement to improve visibility and detail.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local_machine_path/storm_images/image1.jpg"}, "outputs": ["restored_image_path"]}], "source": "benchmark"} {"task_id": 209, "task_desc": "Can you identify and segment all the ships and airplanes in the high-resolution satellite images from the recent disaster-affected region? The images are stored in the folder: /data/disaster_images/. The specific image files are located at: /data/disaster_images/image1.tif. These images contain high-resolution data of the affected areas, capturing detailed views of the landscape and any objects present, such as ships and airplanes.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/disaster_images/image1.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 210, "task_desc": "Can you help me determine the predicted water depths during a flood event using the rainfall data I have? The data is located at /data/rainfall_data.csv. Additionally, you will need the digital elevation model (DEM) data located at /data/region_topography.dem for the ground elevation data.", "structured_plan": [{"agent": "Flood_depth_prediction", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"region_topography_path": "/data/region_topography.dem", "precipitation_data_path": "/data/rainfall_data.csv"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 211, "task_desc": "Can you assess how a sudden closure of major roads due to a natural disaster would affect the availability and demand for alternative transport modes using the data from '/data/road_closure_impact.csv'? I'm interested in understanding the shifts in mobility patterns and the interactions between different transport options during this event. Additionally, the analysis will utilize historical multimodal mobility data and temporal covariates to predict changes in mobility patterns. The historical multimodal mobility data is stored in '/data/historical_mobility_data_tensor.csv', and the temporal covariates are stored in '/data/temporal_covariates_matrix.csv'.", "structured_plan": [{"agent": "Multimodal_mobility_prediction_under_events", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"mobility_tensor_path": "/data/historical_mobility_data_tensor.csv", "temporal_covariates_path": "/data/temporal_covariates_matrix.csv"}, "outputs": ["predicted_mobility_path"]}], "source": "benchmark"} {"task_id": 212, "task_desc": "Can you analyze the satellite image sequences to identify areas affected by flooding and classify the extent of the flood impact? The data is located at /local/data/satellite/images/flood_sequence_01.png, /local/data/satellite/images/flood_sequence_02.png, /local/data/satellite/images/flood_sequence_03.png. These images are temporal sequences of satellite images, each with 3 channels (Red, Green, Blue).", "structured_plan": [{"agent": "Temporal_Image_Sequence_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/satellite/images/flood_sequence_01.png"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/satellite/images/flood_sequence_02.png"}, "outputs": ["classification_results_path"]}, {"agent": "Temporal_Image_Sequence_Classifier", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_sequence_path": "/local/data/satellite/images/flood_sequence_03.png"}, "outputs": ["classification_results_path"]}], "source": "benchmark"} {"task_id": 213, "task_desc": "I have surveillance footage from a wildfire-prone area. Can you analyze these videos to identify any unusual activities or events that could suggest the start or spread of a wildfire? Please provide details on the types of anomalies detected. The video files are located at /local/data/wildfire_footage/video1.mp4, /local/data/wildfire_footage/video2.mp4, and /local/data/wildfire_footage/video3.mp4. These files contain continuous surveillance footage captured over several days, focusing on areas with high wildfire risk.", "structured_plan": [{"agent": "Video_anomaly_detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"extracted_video_feature_path": "/local/data/wildfire_footage/video1.mp4"}, "outputs": ["coarse_grained_anomaly_confidence_path", "fine_grained_anomaly_map_path"]}, {"agent": "Video_anomaly_detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"extracted_video_feature_path": "/local/data/wildfire_footage/video2.mp4"}, "outputs": ["coarse_grained_anomaly_confidence_path", "fine_grained_anomaly_map_path"]}, {"agent": "Video_anomaly_detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"extracted_video_feature_path": "/local/data/wildfire_footage/video3.mp4"}, "outputs": ["coarse_grained_anomaly_confidence_path", "fine_grained_anomaly_map_path"]}], "source": "benchmark"} {"task_id": 214, "task_desc": "I want to assess the impact of a recent flood by comparing pre-flood and post-flood satellite images. The images are stored at '/local/data/pre_flood_images/image1_pre_flood.tif' and '/local/data/post_flood_images/image1_post_flood.tif'. Can you generate a change map that shows the areas affected by the flood?", "structured_plan": [{"agent": "Change_Mapping_and_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_1_path": "/local/data/pre_flood_images/image1_pre_flood.tif", "image_2_path": "/local/data/post_flood_images/image1_post_flood.tif"}, "outputs": ["change_map_path"]}], "source": "benchmark"} {"task_id": 215, "task_desc": "I have low-resolution satellite images showing the development of a flood over a week, and a high-resolution image from the beginning of the flood. Can you help produce a high-resolution image of the flood at a later date using these images? The data is located at /local/data/flood_images/low_res_image_T.jpg for the low-resolution image at time T, /local/data/flood_images/high_res_image_T.jpg for the high-resolution image at time T, /local/data/flood_images/low_res_image_T_prime.jpg for the low-resolution image at time T', and /local/data/flood_images/metadata.json for the metadata associated with the satellite images. The metadata includes latitude, longitude, ground sampling distance (GSD), cloud cover fraction, year, month, and day.", "structured_plan": [{"agent": "Temporal_High_Resolution_Image_Generation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"low_res_image_T_path": "/local/data/flood_images/low_res_image_T.jpg", "high_res_image_T_path": "/local/data/flood_images/high_res_image_T.jpg", "low_res_image_T_prime_path": "/local/data/flood_images/low_res_image_T_prime.jpg", "metadata_path": "/local/data/flood_images/metadata.json"}, "outputs": ["high_res_image_T_prime_path"]}], "source": "benchmark"} {"task_id": 216, "task_desc": "Can you identify the key patterns of population return to urban areas after a disaster using the dataset found at '/data/disaster_recovery/urban_return.csv'? This will help us understand the timeline and scale of population resettlement. Additionally, you will need the following data paths: '/data/disaster_recovery/initial_abnormal_population_mobility_graph.json' for the initial state of abnormal population mobility immediately after the disaster, and '/data/disaster_recovery/normal_population_mobility_graph.json' for the normal state of population mobility prior to the disaster.", "structured_plan": [{"agent": "Post_Disaster_Mobility_Recovery", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"initial_abnormal_population_mobility_graph_path": "/data/disaster_recovery/initial_abnormal_population_mobility_graph.json", "normal_population_mobility_graph_path": "/data/disaster_recovery/normal_population_mobility_graph.json"}, "outputs": ["predicted_population_mobility_graph_path"]}], "source": "benchmark"} {"task_id": 217, "task_desc": "Please analyze the satellite imagery data from the recent storm event to identify any areas that exhibit significant changes in terrain due to landslide activity. The imagery files can be found at '/data/storm_event_imagery/satellite_image_001.tif', '/data/storm_event_imagery/satellite_image_002.tif', and '/data/storm_event_imagery/satellite_image_003.tif'. These files contain high-resolution images capturing the affected areas. Generate an anomaly map to highlight these changes for assessment.", "structured_plan": [{"agent": "Landslide_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/storm_event_imagery/satellite_image_001.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/storm_event_imagery/satellite_image_002.tif"}, "outputs": ["anomaly_map_path"]}, {"agent": "Landslide_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/storm_event_imagery/satellite_image_003.tif"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 218, "task_desc": "Can you analyze the satellite images captured after the wildfire incident and classify the areas affected by the fire into appropriate categories? The images are located at /local/data/wildfire/images/satellite_image.tif. The binary mask array specifying which regions of the satellite image are visible (value 0) or masked (value 1) during the inference process is located at /local/data/wildfire/masks/mask_array.npy.", "structured_plan": [{"agent": "Multi_Spectral_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/wildfire/images/satellite_image.tif", "mask_path": "/local/data/wildfire/masks/mask_array.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 219, "task_desc": "Can you examine the aerial photographs stored at '/data/urban_flooding/aerial_photo_01.jpg', '/data/urban_flooding/aerial_photo_02.jpg', and '/data/urban_flooding/aerial_photo_03.jpg' to identify any unusual patterns or anomalies that could suggest areas of urban flooding after a recent storm? This will assist in understanding the impact and planning for necessary interventions. The data consists of high-resolution satellite images of urban areas captured after the storm.", "structured_plan": [{"agent": "Urban_Anomaly_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/urban_flooding/aerial_photo_01.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/urban_flooding/aerial_photo_02.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Urban_Anomaly_Detection", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/urban_flooding/aerial_photo_03.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 220, "task_desc": "I have a set of social media posts from people affected by a recent hurricane. Can you identify all the place names mentioned in these posts? The posts are located in the directory '/data/hurricane_posts/posts.txt'. The data consists of text files containing social media posts, each file representing a collection of posts from different users.", "structured_plan": [{"agent": "Toponym_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"sentence": "/data/hurricane_posts/posts.txt"}, "outputs": ["detected_toponyms_path"]}], "source": "benchmark"} {"task_id": 221, "task_desc": "Can you examine the radar data from '/local/data/precipitation_patterns.csv' to predict the likelihood of future precipitation and potential impact of flash flooding in the area over the next few hours? This prediction is essential for issuing timely warnings and preparing evacuation plans if necessary. The data file '/local/data/precipitation_patterns.csv' contains time-sequenced radar echo frames representing observed precipitation data over time. You can find the area digital elevation model (DEM) data located at '/local/data/region_topography.dem'.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/local/data/precipitation_patterns.csv"}, "outputs": ["predicted_radar_echo_frames_path"]}, {"agent": "Flood_depth_prediction", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_radar_echo_frames_path"]}, "inputs": {"region_topography_path": "/local/data/region_topography.dem", "precipitation_data_path": "-0-"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 222, "task_desc": "I have a collection of satellite images captured after a wildfire. Can you help determine and categorize the types of vegetation that have been impacted by the wildfire? The specific image used for analysis are '/data/wildfire_impact/image1.tif' and mask '/data/wildfire_impact/mask1.npy'. These images contain detailed information about the affected areas and will be used to assess the impact on vegetation.", "structured_plan": [{"agent": "Multi_Spectral_Classifier", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/wildfire_impact/image1.tif", "mask_path": "/data/wildfire_impact/mask1.npy"}, "outputs": ["reconstructed_image_path", "feature_representation_path"]}], "source": "benchmark"} {"task_id": 223, "task_desc": "Can you identify and count the number of vehicles present in images taken during a nighttime traffic incident? The images are stored in the folder '/data/night_traffic_incidents/'. The specific images to be analyzed are '/data/night_traffic_incidents/image1.jpg'. The image is captured under low-light conditions and require specialized processing to accurately detect vehicles.", "structured_plan": [{"agent": "Low-Light_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/night_traffic_incidents/image1.jpg"}, "outputs": ["detections_path"]}], "source": "benchmark"} {"task_id": 224, "task_desc": "Can you identify and list the locations of all emergency vehicles present in a set of images taken during foggy weather? The images can be found in the folder located at '/local_machine/data/emergency_vehicle_images/image1.jpg'. The image is captured in foggy conditions and contain various emergency vehicles that need to be detected.", "structured_plan": [{"agent": "Foggy_Scenario_Object_Detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local_machine/data/emergency_vehicle_images/image1.jpg"}, "outputs": ["detected_objects_path"]}], "source": "benchmark"} {"task_id": 225, "task_desc": "Can you analyze the satellite images captured before and after the hurricane to determine the extent of damage to residential buildings in the affected city? The images are located at /data/satellite_images/pre_hurricane/image1_pre.png and /data/satellite_images/post_hurricane/image1_post.png. The pre-hurricane image (image1_pre.png) provides a baseline view of the area, while the post-hurricane image (image1_post.png) shows the changes and potential damage caused by the hurricane.", "structured_plan": [{"agent": "Building_damage_assessment", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"pre_disaster_image_path": "/data/satellite_images/pre_hurricane/image1_pre.png", "post_disaster_image_path": "/data/satellite_images/post_hurricane/image1_post.png"}, "outputs": ["damage_classification_map_path", "segmented_buildings_path"]}], "source": "benchmark"} {"task_id": 226, "task_desc": "I have a series of images captured from a city square during a festival, with weather conditions ranging from clear skies to snow. I need to analyze these images to find out the peak crowd size during the event. The images are stored in the directory: /data/festival_images/. The specific image file is: /data/festival_images/image1.jpg", "structured_plan": [{"agent": "Crowd_Counting_in_Adverse_Weather", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/festival_images/image1.jpg"}, "outputs": ["density_map_path", "crowd_count_path"]}], "source": "benchmark"} {"task_id": 227, "task_desc": "Can you analyze the forest images located at '/local/data/forest_images/image1.jpg', '/local/data/forest_images/image2.jpg', and '/local/data/forest_images/image3.jpg' to detect any unusual patterns that might indicate the presence of unauthorized roads or trails? The data consists of high-resolution satellite images of forested areas, which are used to identify any anomalies such as unauthorized roads or trails.", "structured_plan": [{"agent": "Anomaly_Detection_Forest", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/forest_images/image1.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/forest_images/image2.jpg"}, "outputs": ["anomaly_map_path"]}, {"agent": "Anomaly_Detection_Forest", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/forest_images/image3.jpg"}, "outputs": ["anomaly_map_path"]}], "source": "benchmark"} {"task_id": 228, "task_desc": "I have low-resolution satellite images of a region affected by a recent wildfire. Could you generate high-resolution RGB images from these to help assess the extent of the burned areas? The images are located at '/local/data/satellite_images/low_res_image_01.tif'. The metadata providing contextual details for the satellite image is located at '/local/data/satellite_images/metadata_01.json'.", "structured_plan": [{"agent": "High-Resolution_Image_Reconstructor", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/local/data/satellite_images/low_res_image_01.tif", "metadata_path": "/local/data/satellite_images/metadata_01.json"}, "outputs": ["high_resolution_image_path"]}], "source": "benchmark"} {"task_id": 229, "task_desc": "I have a set of images captured during a heavy rainstorm, and they appear to be obscured by rain streaks and poor visibility. Could you help in improving the clarity and detail of these images? The images are located at /local/data/rainstorm_images/image1.jpg, /local/data/rainstorm_images/image2.jpg, and /local/data/rainstorm_images/image3.jpg. These images are affected by rain streaks and reduced visibility, requiring enhancement to restore clarity and detail.", "structured_plan": [{"agent": "Weather_Degraded_Image_Restoration", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/rainstorm_images/image1.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/rainstorm_images/image2.jpg"}, "outputs": ["restored_image_path"]}, {"agent": "Weather_Degraded_Image_Restoration", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"weather_degraded_image_path": "/local/data/rainstorm_images/image3.jpg"}, "outputs": ["restored_image_path"]}], "source": "benchmark"} {"task_id": 230, "task_desc": "Can you assist in segmenting and identifying all the airplanes visible in the high-resolution satellite images captured after the recent hurricane? The images are stored in the directory: /data/hurricane_aftermath_images/. The specific image files to be used are: /data/hurricane_aftermath_images/image1.tif, /data/hurricane_aftermath_images/image2.tif, and /data/hurricane_aftermath_images/image3.tif. These images contain high-resolution data of the affected areas, which will be used for airplane detection and segmentation.", "structured_plan": [{"agent": "Geospatial_Object_Segmentation", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/hurricane_aftermath_images/image1.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 1, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/hurricane_aftermath_images/image2.tif"}, "outputs": ["segmentation_map_image_path"]}, {"agent": "Geospatial_Object_Segmentation", "step": 2, "dependence": [-1], "dependence_content": null, "inputs": {"image_path": "/data/hurricane_aftermath_images/image3.tif"}, "outputs": ["segmentation_map_image_path"]}], "source": "benchmark"} {"task_id": 231, "task_desc": "Can you help me determine the peak water depth during a recent flood event using my rainfall data? The data is located at /data/recent_rainfall.csv. Additionally, you will need the digital elevation model (DEM) data located at /data/region_topography.dem to provide the ground elevation data for the geographical region being simulated.", "structured_plan": [{"agent": "Flood_depth_prediction", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"region_topography_path": "/data/region_topography.dem", "precipitation_data_path": "/data/recent_rainfall.csv"}, "outputs": ["predicted_water_depths_path"]}], "source": "benchmark"} {"task_id": 232, "task_desc": "Can you review the social media posts collected during a recent natural disaster and identify the different types of events being reported, even if there are no prior examples of these specific events in the dataset? The data for this task is located at the following paths: '/data/social_media_posts/natural_disaster_posts.json' for the social media posts collected during the disaster, and '/data/event_detection/prompts.json' for the predefined text prompts with placeholders, triggers are stored at '/triggers/trigger.npy'. The social media posts file contains text data from various platforms, while the prompts file includes templates to assist in event detection.", "structured_plan": [{"agent": "Event_detection", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"context": "/data/social_media_posts/natural_disaster_posts.json", "prompt": "/data/event_detection/prompts.json", "trigger": "/triggers/trigger.npy"}, "outputs": ["event_type_path", "trigger_positions_path", "confidence_scores_path"]}], "source": "benchmark"} {"task_id": 233, "task_desc": "Predict the future precipitation patterns using radar echo data to assess potential flood risks. Utilize the radar echo frames to forecast precipitation intensity over the next 100 minutes, and convert the predicted data into a time-series format suitable for flood depth prediction. This will aid in disaster management by providing insights into possible flooding scenarios. Please provide the path to the radar echo frames data. The radar echo frames data is located at '/local/data/radar_echo_frames/echo_frame_001.dat'.", "structured_plan": [{"agent": "Precipitation_Nowcasting", "step": 0, "dependence": [-1], "dependence_content": null, "inputs": {"radar_echo_frames_path": "/local/data/radar_echo_frames/echo_frame_001.dat"}, "outputs": ["predicted_radar_echo_frames_path"]}, {"agent": "precipitation_data_convert_tool", "step": 1, "dependence": [0], "dependence_content": {"0": ["predicted_radar_echo_frames_path"]}, "inputs": {"radar_echo_frames_path": "-0-"}, "outputs": ["time_series_array_path"]}], "source": "benchmark"}