FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation

03/28/2019
by   Huikai Wu, et al.
12

Modern approaches for semantic segmentation usually employ dilated convolutions in the backbone to extract high-resolution feature maps, which brings heavy computation complexity and memory footprint. To replace the time and memory consuming dilated convolutions, we propose a novel joint upsampling module named Joint Pyramid Upsampling (JPU) by formulating the task of extracting high-resolution feature maps into a joint upsampling problem. With the proposed JPU, our method reduces the computation complexity by more than three times without performance loss. Experiments show that JPU is superior to other upsampling modules, which can be plugged into many existing approaches to reduce computation complexity and improve performance. By replacing dilated convolutions with the proposed JPU module, our method achieves the state-of-the-art performance in Pascal Context dataset (mIoU of 53.13 ADE20K dataset (final score of 0.5584) while running 3 times faster.

READ FULL TEXT

page 2

page 3

page 9

page 11

page 12

page 13

page 14

page 15

research
08/24/2020

EfficientFCN: Holistically-guided Decoding for Semantic Segmentation

Both performance and efficiency are important to semantic segmentation. ...
research
03/19/2018

ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation

We introduce a fast and efficient convolutional neural network, ESPNet, ...
research
07/10/2022

SFNet: Faster, Accurate, and Domain Agnostic Semantic Segmentation via Semantic Flow

In this paper, we focus on exploring effective methods for faster, accur...
research
02/24/2020

Semantic Flow for Fast and Accurate Scene Parsing

In this paper, we focus on effective methods for fast and accurate scene...
research
04/15/2016

High-performance Semantic Segmentation Using Very Deep Fully Convolutional Networks

We propose a method for high-performance semantic image segmentation (or...
research
07/26/2019

Single Level Feature-to-Feature Forecasting with Deformable Convolutions

Future anticipation is of vital importance in autonomous driving and oth...
research
06/30/2017

Superpixel-based Semantic Segmentation Trained by Statistical Process Control

Semantic segmentation, like other fields of computer vision, has seen a ...

Please sign up or login with your details

Forgot password? Click here to reset