Face Recognition Intelligent System Based On Deep Learning

With the development of artificial neural network,deep learning plays an important role in the field of image processing and computer vision.As an important branch of computer vision,face recognition using deep learning has gained more and more attention in society.The intelligent system of face recognition can be divided into two modules from the technical field: face detection and face recognition.Face detection uses a face classifier to detect the presence of a face in each frame of the input video stream,and to identify the presence of a face.Face recognition is to detect the face images detected by down-sampling,normalization and other operations.Using the excellent algorithm to detect the face feature extraction,and then,to compare the feature detected with the facial features in the feature database.If the similarity between the identified features and the features in the database reaches a certain threshold,the recognition results will be output.The main work of this paper is as follows:(1)This paper summarizes the research status of deep learning and face recognition,the advantages and disadvantages of related methods of face detection and face recognition are compared,reading a lot of relevant data and realize the intelligent face recognition system based on deep learning.(2)Use the advanced learning framework—Caffe to train face model files.Considering the lack of BP neural network,we use deep convolution neural network to achieve face recognition.Contrasting several mainstream deep learning frameworks,we use the Caffe framework to train the face model.(3)Build intelligent face recognition system using the trained model file.Design the entire system,making the intelligent system more intelligent,humane.Use MFC with Open CV library programming system interface for the actual needs of attendance work.Test the system in the real scene by changing the camera resolution,face sample database,threshold size,and profile of different angle and improve it.In this paper,the training and test samples of the human face recognition intelligent system are tested by CASIA face database.After the face samples are converted to LMDB format,normalized and other preprocessing operations,we build a network description file and a model configuration file for model training.After parameter adjustment,model optimization and other operations,observe the change curve of the loss value during the training and select the most suitable face feature model.Use the model to design and construct the face recognition intelligent system.The system contains information query,batch registration and other functions.The results show that the system can meet the requirements of the intelligent system of face recognition.We adopt the deep learning algorithm for intelligent face recognition system that has achieved good feasibility and practicability.