Restore Fingerprints Using Pix2Pix

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

1 Scopus citations

Abstract

Previously studied fingerprint readers usually used Minutiae feature. Minutiae uses both directional maps and skeleton images because of its high FMR (False Match Rate). But unlike security attacks on Minutiae, research on directional maps and skeletal image attacks is not going well. In this paper, fingerprint images are generated using the new Pix2Pix model and analyzed representation attack vulnerabilities for the images. When the restored fingerprint by the model was recognized to the fingerprint recognizer, it showed a high recognition success rate and demonstrated the vulnerability for representation attacks on fingerprint readers that also use skeletal images.

Original languageEnglish
Title of host publicationAdvances in Computer Science and Ubiquitous Computing - CSA-CUTE 2019
EditorsJames J. Park, Simon James Fong, Yi Pan, Yunsick Sung
PublisherSpringer Science and Business Media Deutschland GmbH
Pages489-494
Number of pages6
ISBN (Print)9789811593420
DOIs
StatePublished - 2021
Event11th International Conference on Computer Science and its Applications, CSA 2019 and 14th KIPS International Conference on Ubiquitous Information Technologies and Applications, CUTE 2019 - Macao, China
Duration: 18 Dec 201920 Dec 2019

Publication series

NameLecture Notes in Electrical Engineering
Volume715
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

Conference11th International Conference on Computer Science and its Applications, CSA 2019 and 14th KIPS International Conference on Ubiquitous Information Technologies and Applications, CUTE 2019
Country/TerritoryChina
CityMacao
Period18/12/1920/12/19

Keywords

  • Bio-feature
  • Fingerprint
  • Fingerprint reader
  • Pix2Pix
  • Representation attack

Fingerprint

Dive into the research topics of 'Restore Fingerprints Using Pix2Pix'. Together they form a unique fingerprint.

Cite this