The implementation about depth-wise conv

dvlab-research / LargeKernel3D

LargeKernel3D: Scaling up Kernels in 3D Sparse CNNs (CVPR 2023)

https://arxiv.org/abs/2206.10555

Apache License 2.0

197 stars 8 forks source link

The implementation about depth-wise conv #1

Closed FengZicai closed 2 years ago

FengZicai commented 2 years ago

Hi, thanks for your work, I find that depth-wise conv is used. But this feature is not implemented in spconv and MinkowskiNet. Did you implement it yourself? Will the code of depth-wise conv be open?

yukang2017 commented 2 years ago

Yes. We have implemented some attributes by ourselves and combined it into the spconv. We will release both our updated spconv and the codes for this paper in the next month.

Thanks for your interest in our work.

FengZicai commented 2 years ago

I see, thanks for your reply.

YehLi commented 2 years ago

When will the code be released?

yukang2017 commented 2 years ago

Sorry for the late reply. Thanks for your patience. It will be released in one or two weeks.

vehxianfish commented 2 years ago

Sorry for the late reply. Thanks for your patience. It will be released in one or two weeks.

waiting ...

YehLi commented 2 years ago

Sorry for the late reply. Thanks for your patience. It will be released in one or two weeks.

waiting ...

waiting ...

vehxianfish commented 2 years ago

Sorry for the late reply. Thanks for your patience. It will be released in one or two weeks.

waiting ...

waiting ...

waiting

YehLi commented 1 year ago

Sorry for the late reply. Thanks for your patience. It will be released in one or two weeks.

waiting ...

waiting ...

waiting

waiting ...

yukang2017 commented 1 year ago

Sorry for keep you waiting. I have released the core spatial sparse conv in this repo spconv-plus. The depth-wise conv (channel-wise group convolution) can also be found there.