FVeinSyn: Synthetic Finger Vein Image Generator
Authors: Yifan Wang, Jie Gui, Adams Wai Kin Kong, Baosheng Yu, Changsheng Chen, Qi Li, Zhenan Sun, James Tin-Yau Kwok, Alex Kot
Subject: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Submitted: 27 August 2026
Abstract
A major challenge in finger vein recognition is the lack of large-scale public datasets. Existing datasets contain few identities and limited samples per finger, which restricts the advancement of deep learning-based methods. To address this, we propose FVeinSyn, a large-scale, controllable synthetic data generation framework for finger vein images. FVeinSyn explicitly decouples the synthesis of vascular topology from imaging appearance, mitigating the limitations caused by insufficient training dataβsuch as inadequate identity diversity and restricted realism. The framework operates in three stages:
- Vascular Topology Generation: A finger vein identity generator models vascular topology under physiological and geometric constraints using stochastic L-systems, producing anatomically valid and identity-distinctive vascular patterns.
- Realistic Image Rendering: A cascaded region-aware GAN renders the topological maps into realistic near-infrared finger vein images.
- Intra-Class Variation Modeling: An intra-class diversity generator introduces geometric and optical perturbations to simulate realistic variations within the same identity.
Using FVeinSyn, we generated 500,000 images (10,000 vein identities, 50 samples per identity) and conducted extensive evaluations. The results demonstrate that FVeinSyn achieves significant advantages in realism, identity diversity, vascular pattern consistency, and intra-class diversity. Models trained with FVeinSyn outperformed real-data-only baselines across eight public datasets, with an average accuracy improvement of 27.43%.
The code is publicly available at: https://github.com/EvanWang98/Synthetic-Finger-Vein-Generator
Cite as: arXiv:2608.27527 [cs.CV]
via ArXiv CV