Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Submanifold Sparse Convolutional Networks

See original GitHub issue

https://arxiv.org/abs/1706.01307 https://github.com/facebookresearch/SparseConvNet

New paper from Facebook for dealing with sparse, volumetric data. Comes with some nice PyTorch code too! This could be a very useful paper for analyzing binding pockets featurized volumetrically (grid_featurizer does this already).

PRs adding submanifold sparse models to contrib/ welcome!

Issue Analytics

State:
Created 6 years ago
Comments:8 (4 by maintainers)

Top GitHub Comments

1reaction

rbharathcommented, Jul 26, 2017

Sounds good! PRs adding submanifold sparse networks to contrib/ would be very welcome.

0reactions

rbharathcommented, Mar 4, 2018

I’m going to close this issue since it looks like a very involved port, especially since we’ve decided to go all-in on TensorFlow instead of PyTorch for the time being.

Top Results From Across the Web

Submanifold Sparse Convolutional Networks - arXiv

Abstract: Convolutional network are the de-facto standard for analysing spatio-temporal data such as images, videos, 3D shapes, etc.

Submanifold sparse convolutional networks - GitHub

We use the term 'submanifold' to refer to input data that is sparse because it has a lower effective dimension than the space...

3D Semantic Segmentation With Submanifold Sparse ...

We introduce new sparse convolutional operations that are designed to pro- cess spatially-sparse data more efficiently, and use them to develop spatially-sparse ......

Sparse Sabmanifold Convolutions - Medium

The authors of “Submanifold Sparse Convolutional Networks” calls this phenomenon the submanifold expansion problem. Submanifolds refer to ...

Submanifold Convolution Explained | Papers With Code

Submanifold Convolution (SC) is a spatially sparse convolution operation used for tasks with sparse data like semantic segmentation of 3D point clouds.