This is the official PyTorch implementation of the paper "TransFG: A Transformer Architecture for Fine-grained Recognition" (Ju He, Jie-Neng Chen, Shuai Liu, Adam Kortylewski, Cheng Yang, Yutong Bai, Changhu Wang, Alan Yuille).
Same here, weights for nabirds model. It seems kind of wasteful that we all need to train the network when the weights we're looking for already have been computed and is stored somewhere.
Could you upload the weights for trained models? I'm personally looking for the weights of the trained nabirds model. Thanks!