Putting 3D Spatially Sparse Networks on a Diet

12/02/2021
by   Junha Lee, et al.
0

3D neural networks have become prevalent for many 3D vision tasks including object detection, segmentation, registration, and various perception tasks for 3D inputs. However, due to the sparsity and irregularity of 3D data, custom 3D operators or network designs have been the primary focus of 3D research, while the size of networks or efficacy of parameters has been overlooked. In this work, we perform the first comprehensive study on the weight sparsity of spatially sparse 3D convolutional networks and propose a compact weight-sparse and spatially sparse 3D convnet (WS^3-ConvNet) for semantic segmentation and instance segmentation. We employ various network pruning strategies to find compact networks and show our WS^3-ConvNet achieves minimal loss in performance (2.15 compression rate). Finally, we systematically analyze the compression patterns of WS^3-ConvNet and show interesting emerging sparsity patterns common in our compressed networks to further speed up inference.

READ FULL TEXT

page 1

page 8

page 14

page 15

research
06/15/2023

Retrospective: EIE: Efficient Inference Engine on Sparse and Compressed Neural Network

EIE proposed to accelerate pruned and compressed neural networks, exploi...
research
06/15/2022

Deep Neural Network Pruning for Nuclei Instance Segmentation in Hematoxylin Eosin-Stained Histological Images

Recently, pruning deep neural networks (DNNs) has received a lot of atte...
research
11/28/2017

3D Semantic Segmentation with Submanifold Sparse Convolutional Networks

Convolutional networks are the de-facto standard for analyzing spatio-te...
research
09/03/2022

Semantic Segmentation in Learned Compressed Domain

Most machine vision tasks (e.g., semantic segmentation) are based on ima...
research
05/10/2019

EdgeSegNet: A Compact Network for Semantic Segmentation

In this study, we introduce EdgeSegNet, a compact deep convolutional neu...
research
08/29/2019

Smaller Models, Better Generalization

Reducing network complexity has been a major research focus in recent ye...
research
05/27/2018

Compact and Computationally Efficient Representation of Deep Neural Networks

Dot product operations between matrices are at the heart of almost any f...

Please sign up or login with your details

Forgot password? Click here to reset