Tracking Pipeline

• Tracking Pipeline helps you to solve the tracking problem more easily.
• Supported detection algorithms: YOLOv5, YOLOv4, YOLOX, NanoDet, ...
• Supported tracking algorithms: Sort, Deepsort, Motpy, ByteTrack, Norfair.

PipeLine

How to use samples?

YOLOv5

• Download: https://drive.google.com/file/d/1Z3miSUwmKMxbMzucjCrUVvdKpcqiq5_P/view?usp=sharing
• Put at: Tracking_Pipeline/Detection/yolov5/weights

YOLOv4

• Download: https://drive.google.com/file/d/1SgNJYjpc568K0RZPgs7a1RAydbe6kcRQ/view?usp=sharing
• Put at: Tracking_Pipeline/Detection/yolov4/weights

YOLOX

• Download: https://drive.google.com/file/d/1R9SVlocFeInt2Z7fNujYl0CqdqVMX5_c/view?usp=sharing
• Put at: Tracking_Pipeline/Detection/yolox/weights

NanoDet

• Download: https://drive.google.com/file/d/1fupJv-pziAgg0v9dliP2MFqVDdukVqRP/view?usp=sharing
• Put at: Tracking_Pipeline/Detection/nanodet/weights

Deepsort

• Download: https://drive.google.com/file/d/1onB8c0BOO4xhRD1k2erXGz6P-UU-r4Iy/view?usp=sharing
• Put at: Tracking_Pipeline/Tracking/deep_sort/deep/checkpoint

Modify algorithm in Tracking_Pipeline/tracking_config.yaml

References

• https://github.com/Megvii-BaseDetection/YOLOX
• https://github.com/ultralytics/yolov5
• https://github.com/Tianxiaomo/pytorch-YOLOv4
• https://github.com/RangiLyu/nanodet
• https://github.com/mikel-brostrom/Yolov5_DeepSort_Pytorch
• https://github.com/abewley/sort
• https://github.com/ifzhang/ByteTrack
• https://github.com/tryolabs/norfair
• https://github.com/wmuron/motpy