Xin JinHongyu ZhuSimon FongLobo Marques, Joao AlexandreJoao AlexandreLobo MarquesHuafeng QinYun Jiang2026-02-032026-02-032025-02-14https://dspace.usj.edu.mo/handle/123456789/697310.1109/ISCBI64586.2025.11015373As a representative of a new generation of biometrics, vein identification technology offers a high level of security and convenience. Convolutional neural networks (CNNs), a prominent class of deep learning architectures, have been extensively utilized for vein identification. Since insufficient training samples, however, they are unable to extract global feature representations from vein images effectively. To address this issue, we propose StarMixup, a more suitable mixup method for palm vein images since it couls effectively the distribution of vein features to expand samples. Extensive experiments were conducted to validate the performance of StarMixup on two public palm-vein datasets and three general datasets. The results demonstrated that StarMixup provided superior augmentation, and exhibited more stable performance gains compared to mainstream approaches, resulting in the highest recognition accuracy and lowest identification error.entext::conference output::conference proceedings