[fix] integer convolution and unsplit #18
Merged
Conversation
This file contains hidden or 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
Collaborator
|
integer_conv_weight should be False as a defulat behaviour for now. |
Contributor
Author
|
The kmac/pixel calculation for PANDASET needs to be improved. |
sarhakor
changed the title
Collaborator
Please check later once my chages are merged. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
This is a PR for bug fixes related to the following MR in FCTM:
https://git.mpeg.expert/MPEG/Video/fcm/fctm/-/merge_requests/81#8ec9a00bfd09b3190ac6b22251dbb1aa95a0579d
There appear to be several items that need to be addressed on the CompressAI-Vision side:
It appears that the unsplit performance evaluation code is missing support for Pandaset.
(added)Currently, all integer_conv_weight options are set to False by default in compressai_vision/cfgs/vision_model/default.yaml. Therefore, I have changed them to True.
(changed)When running panoptic segmentation, the integer_conv_weight key seems to be missing, for example, in compressai_vision/cfgs/vision_model/default.yaml. This key has been added for the function to work properly.
(added)When setting integer_conv_weight to True for panoptic segmentation, integer convolution does not seem to be working correctly. (There seems to be a bug.)
(FCTM side bug)Regarding the class-wise calculation, it seems that “Cactus” is still included.
(required)The kmac/pixels calculation code does not include PANDASET.
(required)