Real-Time Object Detection & Tracking
Video detection pipeline with YOLO and CNN classifiers; benchmarked YOLOv5, YOLOv8 and DETR on custom datasets.
This project combines YOLO detection with CNN classifiers in a video-processing pipeline. YOLOv5, YOLOv8 and DETR were benchmarked on custom datasets to compare detection approaches. PyTorch, OpenCV and TensorFlow form the implementation stack.
Technologies
PyTorch, OpenCV, TensorFlow
How it works
- Video frames
- Object detection
- CNN classification
- Tracking output
