ChickEgg is a real-time object detection system built using YOLOv11, designed to accurately identify chickens and eggs in farm environments from live video streams.
- β‘ Real-time detection with YOLOv11
- π― High performance on a custom farm dataset:
- Precision (B): 0.93
- Recall (B): 0.92
- [email protected] (B): 0.945
- [email protected]:0.95 (B): 0.735
- πΈ Custom data collection and annotation using Roboflow
- π Trained specifically on farm-related scenes for detecting:
chicken
egg
- Total annotated images: 6,715
- Training set: 5,524 images
- Validation set: 792 images
- Test set: 399 images
- Collected from various farm camera feeds
- Annotated using Roboflow for high-quality bounding boxes and labels
- Balanced to ensure robust generalization across unseen farm environments
Feel free to contribute, raise issues, or fork this project to extend its capabilities!