Generalized Sparse Convolution #77
Merged
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
CCInc
reviewed
Jun 8, 2021
There was a problem hiding this comment.
Great work! I will take a closer look more this week and test it.
I was wondering, will this make cpp functions like convert_map_forward and insertion_forward obsolete? Unless maybe somebody is using torchsparse.nn.functional.voxelize outside of torchsparse.nn.functional.downsample for some reason? We might want to see if we can remove these
zhijian-liu
reviewed
Jun 10, 2021
CCInc
reviewed
Jun 11, 2021
CCInc
reviewed
Jun 11, 2021
CCInc
previously approved these changes
Jun 14, 2021
zhijian-liu
previously approved these changes
Jun 21, 2021
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Outdated suggestions cannot be applied.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.
In this PR we implement the generalized sparse convolution operation. This operation mainly differs from the current sparse convolution in downsampling stage with
kernel_sizenot equal tostride. The output will be dilated (similar to stridednn.Conv3d) and is not calculated withc_out = torch.floor(c_in // 2) * 2. Besides, we also added support for asymmetric convolutional kernel sizes and strides.