utku topcuoglu

Thank you for your contribution. In the line below, the total iterations per epoch is assumed to be 1000 (regardless of the batch size or dataset). It is not clear...

Sinkhorn distance implementation

1

Thanks for the great contribution! As mentioned in https://github.com/Megvii-BaseDetection/OTA/issues/17#issue-1484922452, the original sinkhorn implementation is different (page 11 of your paper):  Why is this the case?

Instance Segmentation masks are all 0 images

3

Thanks for your contribution. Instance segmentation masks provided for your dataset (in the drive link) are all only 0-valued images (no mask available). Also, the links for instance, and normal...

I want to be sure about the learning rates you have used for fine-tuning. In the paper you state:  And, in this repo, it is written that for 100...