ClickI Love Computer Vision to star and get the latest CVML technologies faster
Today we introduce an open-source library for multi-object tracking implemented in PyTorch. Search for MOT tracking on Github, and this code ranks first, currently having 505 stars.
This code implements the well-known Deep SORT multi-object tracking algorithm. The original implementation by the algorithm’s author is based on TensorFlow, and the author has implemented the RE-ID (Re-Identification) module in PyTorch, replacing the object detection model from Faster RCNN to YOLOv3.
The author provides pre-trained models, but if you want to train your own RE-ID model, that is also supported.
Example of running results:
The author’s results in the video:
The effect looks quite good, and everyone can try it out!
Code address:
https://github.com/ZQPei/deep_sort_pytorch
Multi-object Tracking Discussion Group
Follow the latest cutting-edge single and multi-object tracking technologies, scan the QR code to add CV Jun to join the group (if you are already friends with CV Jun’s other accounts, please message directly).
(Please be sure to indicate: Tracking)
For those who like to communicate on QQ, you can join the official QQ group: 805388940.
(I may not be online all the time, so please excuse if I can’t verify you promptly.)
Long press to followI Love Computer Vision