Skip to main navigation Skip to search Skip to main content

A Review on Palmprint Image-Level Attacks

  • Qiuli Zhang
  • , Kaihua Zheng
  • , Jinhui Xu
  • , Yong Xu
  • , Jinrong Cui*
  • *Corresponding author for this work

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

Abstract

With the popularization of biometric recognition technology, palmprint recognition has attracted widespread attention because of its high uniqueness and non-replicability. However, recent studies have shown that various image-level attacks on palmprint recognition systems can interfere with the decision-making process of the model, leading to a decline in both the accuracy and security of the system. In this paper, we first summarize the applications of palmprint recognition technology and then describe two main types of image-level attacks: adversarial attacks and reconstruction attacks. Finally, we summarize existing works on image-level attacks targeting palmprint recognition, highlighting key methods and strategies.

Original languageEnglish
Title of host publicationBiometric Recognition - 19th Chinese Conference, CCBR 2025, Proceedings
EditorsWei Jia, Lu Leng, Weidong Min, Jun Chu, Jie Gui, Xiangbo Shu, Xianye Ben, Zhenan Sun, Yuming Fang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages122-130
Number of pages9
ISBN (Print)9789819561223
DOIs
StatePublished - 2026
Externally publishedYes
Event19th Chinese Conference on Biometric Recognition, CCBR 2025 - Nanchang, China
Duration: 21 Nov 202523 Nov 2025

Publication series

NameLecture Notes in Computer Science
Volume16360 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference19th Chinese Conference on Biometric Recognition, CCBR 2025
Country/TerritoryChina
CityNanchang
Period21/11/2523/11/25

Keywords

  • Adversarial Attack
  • Palmprint Recognition
  • Reconstruction Attack
  • Security

Fingerprint

Dive into the research topics of 'A Review on Palmprint Image-Level Attacks'. Together they form a unique fingerprint.

Cite this