Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,64 @@
+---
+license: mit
+tags:
+- object-detection
+- yolov8
+- pothole-detection
+- road-repair-verification
+- computer-vision
+- india
+datasets:
+- roboflow/indian-pothole-dataset
+- custom
+language:
+- en
+---
+
+# QuickFix It — YOLOv8 Pothole Detection Model
+
+Fine-tuned YOLOv8n model for detecting and verifying pothole repairs on Indian roads.
+
+## Model Details
+
+- **Architecture**: YOLOv8n (nano)
+- **Task**: Pothole detection & repair verification
+- **Training Images**: 2,281
+  - ~1,243 from Roboflow Indian Pothole Dataset
+  - 1,002 from 501 custom before/after road repair pairs
+- **Epochs**: 25
+- **Model Size**: ~6.2 MB
+
+## Usage
+
+```python
+from ultralytics import YOLO
+from huggingface_hub import hf_hub_download
+
+# Download model
+model_path = hf_hub_download(
+    repo_id="RoxieRoller/QuickFixIt-model",
+    filename="best.pt"
+)
+
+# Load and run inference
+model = YOLO(model_path)
+results = model("your_road_image.jpg")
+results[0].show()
+```
+
+## About QuickFix It
+
+This model is part of the **QuickFix It** civic tech platform — an AI-powered pothole repair verification system for Indian municipalities.
+
+GitHub: [https://github.com/mohammedhussain06/QuickFix](https://github.com/mohammedhussain06/QuickFix)
+
+The model is used in a 5-step verification pipeline:
+1. Image Integrity Check
+2. GPS Consistency Check
+3. Camera Angle Analysis
+4. Landmark Matching
+5. **Repair Detection (this model)**
+
+## License
+
+MIT