计算机工程与应用2023，Vol.59Issue(24)：98-109,12.DOI:10.3778/j.issn.1002-8331.2208-0086
融合Transformer和CNN的手掌静脉识别网络
Palm Vein Recognition Network Combining Transformer and CNN
摘要
Abstract
Aiming at the low accuracy of palm vein feature extraction and recognition,it proposes a palm vein recogni-tion network PVCodeNet.It designs an improved BasicBlock and Transformer Encoder,and uses AAM-loss(additional angular margin loss)to expand decision boundary.It successfully applies Transformer Encoder to global feature extrac-tion of palm vein firstly.Improved BasicBlock uses Do-Conv to replace Conv for feature extraction,it makes extracted features more distinctive.it also adds standardized attention module NAM,its detailed features of in channel and spatial domain are extracted by applying heavy sparsity penalty to suppress weights of insignificant features.This paper describes in detail the key point location,ROI extraction and image enhancement,then makes detailed experiments on feature vec-tor dimension and AAM-loss parameter setting.Finally,ablation experiments are carried out on PolyU database and self-built database SEPAD-PV,EER reaches 0,it achieves a breakthrough in the highest recognition rate.In order to verify the generalization performance of network,it is also verifies on the palmprint database Tongji and the finger vein database SDUMLA with similar texture features.EER is far superior to other mainstream algorithms,which fully proves the superi-ority of this algorithm.关键词
手掌静脉识别/Transformer编码模块/深度超参数化卷积(Do-Conv)/规一化注意力机制(NAM)/扩大决策边界的损失函数(AAM-Loss)Key words
palm vein recognition/Transformer encoder/Do-Conv/normalization-based attention module(NAM)/additive angular margin loss(AAM-Loss)分类
信息技术与安全科学引用本文复制引用
吴凯,沈文忠,贾丁丁,梁娟..融合Transformer和CNN的手掌静脉识别网络[J].计算机工程与应用,2023,59(24):98-109,12.基金项目
国家自然科学基金(61802250). （61802250）
