Few-Shot Classification with Cross-Feature Fusion

Guohui Yao, Min Li*, Dawei Song

*Corresponding author for this work

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

Abstract

Most existing methods for few-shot image classification treat the support features and query features separately, and the category of a query image is determined based on its similarity to the support images. However, the relationships between the query features and support features are often ignored. In this paper, we propose a strategy that exploits the relationships between query and support features and give more weights to the highly related features. A cross-feature fusion method is further proposed to combine with metric learning to reduce overfitting. Extensive experiments on a wide range of datasets show that our method has achieved advanced results.

Original languageEnglish
Title of host publication2023 8th International Conference on Computer and Communication Systems, ICCCS 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages944-949
Number of pages6
ISBN (Electronic)9781665456128
DOIs
Publication statusPublished - 2023
Event8th International Conference on Computer and Communication Systems, ICCCS 2023 - Hybrid, Guangzhou, China
Duration: 21 Apr 202324 Apr 2023

Publication series

Name2023 8th International Conference on Computer and Communication Systems, ICCCS 2023

Conference

Conference8th International Conference on Computer and Communication Systems, ICCCS 2023
Country/TerritoryChina
CityHybrid, Guangzhou
Period21/04/2324/04/23

Keywords

  • Few-shot classificaiton
  • cross-feature
  • query feature
  • similarity
  • support feature

Fingerprint

Dive into the research topics of 'Few-Shot Classification with Cross-Feature Fusion'. Together they form a unique fingerprint.

Cite this