ML / Done
Prototype Chicken Detection Model
Overview
Prototype poultry-monitoring system that detects and counts chickens in images and video. Built with Python, Detectron2, and PyTorch around a RetinaNet model with a ResNet-50 FPN backbone, compressed with dynamic quantization and structured pruning.
Challenge
Developing a high-sensitivity deep learning model for accurate real-time chicken detection in poultry farm environments.
Result
Developed a prototype capable of reliable chicken detection and counting in images and videos, designed for future integration into a real-time monitoring application.
Key Statistics
~150 MB
Model Size
70%
Detection Accuracy
Poultry Farm Monitoring
Target Application
Technologies
Frameworks
-
Detectron2 -
PyTorch
Programming
-
Python
Libraries
-
OpenCV -
NumPy -
Matplotlib
Model Architecture
- RetinaNet with ResNet-50 FPN
Gallery
Chicken Detection Output
Image with detected chickens and bounding boxes.
Video Detection Frame
Frame from video showing result for chicken detection.
