fangruizhu / self_sup_semivos Goto Github PK

I find your work excellent and eager to learn from your implementation of momentum update and other details.Hope you can release the code soon. >_<

I look forward to seeing your implementation!

Hi, I have a question about the online adaptation. During inference, given a video sequence, do you start the online process once you obtain the second frame's predict mask, or wait for all frames' masks in that sequence are produced? If I don't make this clear, please let me know. Looking forward to your reply!

Hello, thanks for your fantastic work.
When implementing MOCO in your paper, the reference frame and the current frame use different feature extractors. Are their parameter initializations the same? Will the parameters of the two feature extractors be much different after the training? What's more, is there anything you should pay attention to when using MOCO?

Thank you very much!

Hello:
thanks for your fantastic work .i have a question about that Why the θr is updated with momentum?

thank you very much!

Hi, thank you for sharing such a great job!
I wonder when will you release the code?
I have some questions about the implementation of the momentum memory. While training, (1) how many previous frames do you use to serve as reference frames? (2) do you use the reference frames to reconstruct the query frame seperately (i.e. one by one) or at the same time (apply a softmax afterwards) like [7]?
I would appreciate it if you could help me. Hope you can release your code soon. Thank you!