基于时序光流与微表情的人脸活体识别
FACE RECOGNITION IN VIVO BASED ON TEMPORAL OPTICAL FLOW AND MICRO-EXPRESSION
周延森 1徐传凯 1崔见泉1
作者信息
- 1. 国际关系学院网络空间安全学院 北京 100091
- 折叠
摘要
人脸活体检测模型存在着泛化性较差、复杂度高等问题,从而导致不能有效识别新假体攻击类型.基于此,该文提出一种基于时序光流和微表情人脸活体检测模型(FT-CNN).该模型由TVNet-DTSCNN和Attention CNN-LSTM卷积网络组成.TVNet-DTSCNN 对输入的时序人脸帧分别进行光流预测和微表情提取,Attention CNN-LSTM提取人脸视频中的运动细节线索并放大,使模型学习到活体和假体人脸的鲁棒性特征.在CASIA、CASIA-SURF和MSU-MFSD数据集上的训练和测试结果表明,FT-CNN在准确率(Acc)、平均错误率(HTER)和泛化性上的表现相比之前的模型均显著提升.
Abstract
Insufficient generalization and complexity in face anti-spoofing detection models results in a poor performance targeting on new face attack algorithm.Therefore,a face recognition model in vivo(FT-CNN)is proposed based on optical flow estimate and micro-expression in face.The model consisted of TVNet-DTSCNN and Attention CNN-LSTM.TVNet-DTSCNN performed optical flow prediction and micro-expression extraction on the input time-series face frames,and attention CNN-LSTM extracted and magnified the motion detail cues in the face video,which made the model to learn the robust feature for both live and prosthetic faces.Experiments on CASIA,CASIA-SURF and MSU-MFSD datasets indicate that the performance of FT-CNN in accuracy(Acc),average error rate(HTER)and generalization is significantly improved compared with the previous models.
关键词
人脸活体检测/微表情识别/注意力机制/3D卷积网络/光流预测Key words
Face anti-spoofing detection/Mirco-expression recognition/Attention mechanism/3D CNN/Optical flow estimate引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据