RedRAINXXXX

Results 3 issues of RedRAINXXXX

You use square patches during training, but how do you handle non-square images during inference? Directly resize non-square images to (224, 224)? Looking forward to your reply;)