摘要
基于MTCNN的思想设计了一种基于视频的人脸检测识别的方法.首先使用MTCNN对视频中的人脸进行检测,然后构建一个CNN网络对人脸进行分类识别,最后和其他常用的CNN网络模型进行对比.实验结果表明本方法在视频人脸识别的性能方面,有较好的表现,人脸检测及识别的准确率较高.
A method of video face detection and recognition was designed based on the idea of MTCNN. In this paper,MTCNN algorithm was used to detect faces based on video,and then a CNN network to recognize faces was constructed,which was finally compared with other CNN network models. The experimental results showed that this method has a good accuracy for video face detection and recognition.
引文
[1]江伟坚,郭躬德,赖智铭.基于新Haar-like特征的Adaboost人脸检测算法[J].山东大学学报:工学版,2014,44(2):43-48.
[2]Kivinen J,Szepesvári,Csaba,et al.On the Expressive Power of Deep Architectures[C].International Conference on Algorithmic Learning Theory:Springer-Verlag,2011,6 925:18-36.
[3]王玮,闵卫东,樊梦丹,等.基于择优检测和多尺度匹配的实时人脸识别[J].计算机工程与设计,2018,39(9):2 957-2 960.
[4]Zhang K,Zhang Z,Li Z,et al.Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks[J].IEEE Signal Processing Letters,2016,23(10):1 499-1 503.
[5]Niu X X,Suen C Y.A novel hybrid CNN-SVM classifier for recognizing handwritten digits[J].Pattern Recognition,2012,45(4):1 318-1 325.
[6]Karen Simonyan,Andrew Zisserman.Very Deep Convolutional Networks for Large-Scale Image Recognition[EB/OL].2015,https://arxiv.org/abs/1502.03167.
[7]Ioffe S,Szegedy C.Batch normalization:accelerating deep network training by reducing internal covariate shift[J].ICML'15,2015,37:448-456.
[8]黄志建.基于证据K近邻人脸识别的研究与系统设计[D].广州:华南理工大学,2018.