KimSoybean

Thank you! I want to apply DCR-v1 on one-stage detector. Do you have some ideas about that?

Thank you！@bowenc0221

In sampling strategies, some pad_indexes are randomly sampled in all boxes or null ( pad_indexes=[] ). This causes there are same samples in positive and negative samples. Do i misunderstand...

I see. Thank you. @bowenc0221

Why you multiply two scores? This makes the original score become lower or do I misunderstand that?

Thanks! @bowenc0221. Have you trained DCR for more than 9 epochs ? Will the over-fitting happen? I want to know more details because I'm using DCR for one-stage face detection.

location of FT: https://github.com/DTennant/CL-Visualizing-Feature-Transformation/blob/62c273726c03a3cdad7aa6b95508d12f583db3cb/memory/mem_moco.py#L122

Further studies on simsiam is performing now, sorry we cannot provide the codes now. You can add the code on simsiam by youself.