update simmim using conv decoder which replace simple header #216
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
oh hey! yea I believe it is equivalent at least, I could do an extra rearrange on the predicted pixels to get back the reconstructed image |
|
Thank you for your reply. Yes, it is. I thought the original restoration method should be followed (using Conv+PixelShuffle), if you don't think it's important then you could close the PR . |
lucidrains
force-pushed
the
main
branch
4 times, most recently
from
70284c0
to
4ef72fc
Compare
|
@wnma3mz i think being able to get back the reconstructed image is a good idea, let me get the function out when i find some time. feel free to leave this open! |
|
It's hard not to support this and it will help many more people, including me. If there is anything I can do to help, please contact me directly. |
lucidrains
force-pushed
the
main
branch
2 times, most recently
from
ddff7a7
to
b3e90a2
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
When I read the code of the authors of the paper, I found that the authors used Conv and PixelShuffle layers. To ensure the effectiveness of the MIM method, maybe we should use this decoder method.
P.S. Unfortunately, I have not modified the loss function for the time being because of the difference in the way the mask is generated.
https://github.com/microsoft/SimMIM/blob/bec329f54c2f84db16974035c3f010a88f1b3eb0/models/simmim.py#L104-L121