ParseNet: Looking Wider to See Better

Wei Liu, Andrew Rabinovich, Alexander C. Berg
6/15/2015
cs.CV

Official code: weiliu89/caffe (4,808 stars, last commit 2023-04-21).

Abstract

We present a technique for adding global context to deep convolutional networks for semantic segmentation. The approach is simple, using the average feature for a layer to augment the features at each location. In addition, we study several idiosyncrasies of training, significantly increasing the performance of baseline networks (e.g. from FCN). When we add our proposed global feature, and a technique for learning normalization parameters, accuracy increases consistently even over our improved versions of the baselines. Our proposed approach, ParseNet, achieves state-of-the-art performance on SiftFlow and PASCAL-Context with small additional computational cost over baselines, and near current state-of-the-art performance on PASCAL VOC 2012 semantic segmentation with a simple approach. Code is available at https://github.com/weiliu89/caffe/tree/fcn .

View on arXivView PDF

Code Implementations(1)

weiliu89/caffeOfficial100%

Caffe: a fast open framework for deep learning.

4,8081,643M, C++Mar 31, 20153 years agoNOASSERTION
deep-learningobject-detectionreal-timesemantic-segmentation

Similar Papers

FCN+: Global Receptive Convolution Makes FCN Great Again

88% match1 repo12023

Efficient Semantic Segmentation via Lightweight Multiple-Information Interaction Network

86% match2024

Context Encoding for Semantic Segmentation

86% match1 repo2.0k2018

Pixel Difference Convolutional Network for RGB-D Semantic Segmentation

86% match2023

SegNeXt: Rethinking Convolutional Attention Design for Semantic Segmentation

85% match2 repos9.9k2022

Understanding Convolution for Semantic Segmentation

85% match1 repo6102017

Cite this paper

@article{liu2015parsenet,
  title  = {ParseNet: Looking Wider to See Better},
  author = {Wei Liu and Andrew Rabinovich and Alexander C. Berg},
  year   = {2015},
  eprint = {1506.04579},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1506.04579}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1506.04579.svg)](https://codeofpaper.com/paper/1506.04579)

Discussion