Face Recognition for Intelligent Robot Safety Verification System
Xingqian Li, Haoyu Zhao, Hongwei Zhao, Jianjun Wang, Peijun Xia · 2017
The current intelligent robots lack a certain degree of security, can not accurately identify the user. Embedded devices exist in the existing face recognition is not accurate enough, the depth of learning resources and other issues is not enough. We presented a new type of ResNet face recognition model for transplanting in embedded devices. We used the machine to learn to face the face detection, and extracted 68 face key points, the use of depth residual network algorithm for further identification, and ultimately compared the European distance to determine and identify the user. Experiments show that the scheme has the advantages of good precision and good accuracy, and can be used in the intelligent robot system.