about separate Batch All (BA+) triplet loss
Are separate Batch All (BA+) triplet loss like gaitset?Gaitset calculates triplet losses independently of n column feature vectors, and then takes the average of n losses as the final loss function. Do you also do that?Can you tell me how you do it?
Thanks for your attention! Yeah, the Batch All triplet loss we utilized should be no different with that employed in the excellent GaitSet.
Hi, the OpenGait is released now! ( https://github.com/ShiqiYu/OpenGait ) This project not only contains the full code of gaitpart but also reproduces several SOTA models of gait recognition. Enjoy it and any questions or suggestions are welcome!