Adversarial Masking Contrastive Learning for vein recognition
Huafeng Qin, Yiquan Wu, Mounim A. El-Yacoubi, Jun Wang, Guangxiang Yang
TL;DR
This paper tackles the data-scarce problem of vein recognition by proposing Adversarial Masking Contrastive Learning (AMCL), which jointly trains a GAN-based mask generator and a contrastive encoder in a min–max framework to produce challenging masked samples that improve robustness. A GAN learns mask distributions from a large set of generated masks, and a contrastive learning objective (SimCLR-style) learns invariant vein representations from masked and unmasked views, with an adversarial term encouraging harder samples via latent-variable optimization. The approach yields a robust encoder that, when combined with a Softmax classifier, achieves state-of-the-art recognition accuracy and low equal error rates on three public palm-vein datasets, outperforming standard contrastive methods (SimCLR, VICReg, ADIOS) and several vein classifiers. Overall, AMCL offers a practical, data-efficient pathway to strong vein recognition performance with potential benefits for privacy-preserving biometric systems.
Abstract
Vein recognition has received increasing attention due to its high security and privacy. Recently, deep neural networks such as Convolutional neural networks (CNN) and Transformers have been introduced for vein recognition and achieved state-of-the-art performance. Despite the recent advances, however, existing solutions for finger-vein feature extraction are still not optimal due to scarce training image samples. To overcome this problem, in this paper, we propose an adversarial masking contrastive learning (AMCL) approach, that generates challenging samples to train a more robust contrastive learning model for the downstream palm-vein recognition task, by alternatively optimizing the encoder in the contrastive learning model and a set of latent variables. First, a huge number of masks are generated to train a robust generative adversarial network (GAN). The trained generator transforms a latent variable from the latent variable space into a mask space. Then, we combine the trained generator with a contrastive learning model to obtain our AMCL, where the generator produces challenging masking images to increase the contrastive loss and the contrastive learning model is trained based on the harder images to learn a more robust feature representation. After training, the trained encoder in the contrastive learning model is combined with a classification layer to build a classifier, which is further fine-tuned on labeled training data for vein recognition. The experimental results on three databases demonstrate that our approach outperforms existing contrastive learning approaches in terms of improving identification accuracy of vein classifiers and achieves state-of-the-art recognition results.