With the rapid development of artificial intelligence

face recognition technology has been widely used in the social and industrial fields. But the traditional face recognition technology has low recognition rate

slow recognition speed and very high environmental requirements. Therefore

it is urgent to innovate the method. This paper aims to study how to deep learning algorithm in face recognition field

by constructing a double heterogeneous depth of the neural network model

simulation of neural network learning

the use of CNN and DBN and many other models make the computer gradually learn to recognize image features based on a large number of data and human face

and in-depth study of the key technologies of face recognition

and the recognition rate to improve the robustness of face recognition technology.