Image Segmentation
ultralytics
PyTorch
yolo
yolo11
deep-learning
satellite
rso-detection
custom_code
Eval Results (legacy)
Instructions to use rayh/astro-seg with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ultralytics
How to use rayh/astro-seg with ultralytics:
from huggingface_hub import hf_hub_download from ultralytics import YOLO # pick the weights file from this repo's "Files and versions" tab weights = hf_hub_download("rayh/astro-seg", "<weights>.pt") model = YOLO(weights) source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True) - Notebooks
- Google Colab
- Kaggle
best
Model Information
This is a YOLO11-based segmentation model for detecting Resident Space Objects (RSOs) in satellite imagery.
Classes
- streak: Class 0
Usage
from huggingface_hub import InferenceClient
client = InferenceClient(model="best")
result = client.image_segmentation(image)
Training Metrics
- mAP@50: 0.8750
- mAP@50-95: 0.6194
- Downloads last month
- -
Evaluation results
- Mean Average Precision (mAP@50) on RSO Detection Datasetself-reported0.875
- Mean Average Precision (mAP@50-95) on RSO Detection Datasetself-reported0.619