Multi-scale Spatially-Asymmetric Recalibration for Image Classification

  • Yan Wang
  • , Lingxi Xie
  • , Siyuan Qiao
  • , Ya Zhang*
  • , Wenjun Zhang
  • , Alan L. Yuille
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Convolution is spatially-symmetric, i.e., the visual features are independent of its position in the image, which limits its ability to utilize contextual cues for visual recognition. This paper addresses this issue by introducing a recalibration process, which refers to the surrounding region of each neuron, computes an importance value and multiplies it to the original neural response. Our approach is named multi-scale spatially-asymmetric recalibration (MS-SAR), which extracts visual cues from surrounding regions at multiple scales, and designs a weighting scheme which is asymmetric in the spatial domain. MS-SAR is implemented in an efficient way, so that only small fractions of extra parameters and computations are required. We apply MS-SAR to several popular building blocks, including the residual block and the densely-connected block, and demonstrate its superior performance in both CIFAR and ILSVRC2012 classification tasks.

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
EditorsVittorio Ferrari, Cristian Sminchisescu, Yair Weiss, Martial Hebert
PublisherSpringer Verlag
Pages523-539
Number of pages17
ISBN (Print)9783030012601
DOIs
StatePublished - 2018
Externally publishedYes
Event15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany
Duration: 8 Sep 201814 Sep 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11217 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference15th European Conference on Computer Vision, ECCV 2018
Country/TerritoryGermany
CityMunich
Period8/09/1814/09/18

Keywords

  • Convolutional Neural Networks
  • Large-scale image classification
  • Multi-Scale Spatially Asymmetric Recalibration

Fingerprint

Dive into the research topics of 'Multi-scale Spatially-Asymmetric Recalibration for Image Classification'. Together they form a unique fingerprint.

Cite this