Closed jiaweihe1996 closed 2 years ago
Oh, it's probably a mistake we made.
In fact, we've read your excellent work, but have not paid attention to the details. We'll revisit this matter, and consider citing your work in the future version of our paper.
Thanks for your reminder, and sorry for the trouble we made. Best wishes.
Hi. There is a related paper Learnable Graph Matching: Incorporating Graph Partitioning with Deep Feature Learning for Multiple Object Tracking, published in CVPR2021. In this paper, we modify DeepSORT as our baseline method, including vanilla matching, strong ReID model, ECC, and EMA for updating appearance features, just like some parts in StongSORT. So, please consider citing GMTracker. Thank you very much.