CVPR Findings2026
Fingerprint Fragment Expansion using Image Outpainting Approach based on Spectral Normalization PatchGAN
C. Zaghetto, A. Purim, W. Oliveira, J. R. Ribeiro, H. Nolla, F. Santos, M. Chang, R. H. Vareto
Abstract
Partial fingerprint authentication systems, such as the ones used in smartphones or tablets, are used for a wide range of tasks. However, the small sensor sizes in these devices limit system performance, creating a trade-off between high false acceptance rates (low security) and high false rejection rates (low usability), or rendering some algorithms unusable. This paper introduces a technique for expanding fingerprint fragments using a Generative Adversarial Network (GAN) model with a patch-based loss function and Spectral Normalization (SN) for weight normalization. The model expands the image by generating ridge lines with local continuity based on neighboring information and global patterns. Our experiments show that expanding fragments with the proposed method enables the reconstruction of minutiae outside the fragment border, reducing the equal error rate (EER) by approximately 0.524% and 0.34% in the NIST Biometric Image Software Suite Bozorth3 and Ver-iFinger 13.1 SDK matchers, respectively.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
Builds on1
Related papers
- Think Twice Before Detecting GAN-generated Fake Images from their Spectral Domain ImprintsChengdong Dong, Ajay Kumar, Eryun LiuCVPR 2022 · 61 citations
- Reality Transform Adversarial Generators for Image Splicing Forgery Detection and LocalizationXiuli Bi, Zhipeng Zhang, Bin XiaoICCV 2021 · 30 citations
- PU-GAN: A Point Cloud Upsampling Adversarial NetworkRuihui Li, Xianzhi Li, Chi-Wing Fu, Daniel Cohen-Or et al.ICCV 2019 · 496 citations
- AFace: Range-flexible Anti-spoofing Face Authentication via Smartphone Acoustic SensingZhaopeng Xu, Tong Liu, Ruobing Jiang, Pengfei Hu et al.UbiComp 2024 · 24 citations
- FakePolisher: Making DeepFakes More Detection-Evasive by Shallow ReconstructionYihao Huang, Felix Juefei-Xu, Run Wang, Qing Guo et al.ACM MM 2020 · 72 citations
