Research On Face Recognition Technology Based On Deep Learning

Identity technology has a wide range of applications for many scenarios.The identity verification through face information is direct and convenient,not easy to be lost,it is easy to accepted by users.It has been paid attention to and extensive research.In recent years,with the research and application of deep learning technology in the field of artificial intelligence such as computer vision,face recognition has been greatly developed as an important part of computer vision technology.There are many new face recognition algorithms that based on deep learning to appear,and those method's effect are much higher than the traditional method,so that this technology with broad prospects can be applied to practical work life.A real-time face recognition system based on deep learning technology has application prospects.This paper focuses on the design and implementation of a face recognition program based on deep learning technology.This paper introduces the development and application of face recognition technology in recent years,and the influence of deep learning on this technology field.The feasibility analysis and demand analysis of the program were carried out.This paper proposes the functions of the program,then analyzes the workflow of the program.According to the workflow,the program was decomposed into six functional modules as image acquisition,face detection,feature extraction,identity recognition,emotional recognition and result outputting.After the article,the design and implementation methods of the main functional modules are introduced.The program performs face detection through the deep learning model YOLO.The face alignment is performed by the ERT algorithm,the 128-dimensional feature vector is extracted by ResNet,and on this basis,the task of face recognition is realized.The input and output are realized by OpenCV.At the end of the paper,the program function was tested.On the flw dataset,the program can achieve an accuracy of more than 98%.Deep learning and neural network technology are very suitable for image recognition,especially face recognition.The face recognition program developed based on this technology with the support of appropriate hardware can achieve higher recognition accuracy and faster operation,fulfil requirements the requirements of practical applications.