Emotional Neural Network And Its Application In Face Recognition Research

The twenty-first century is the age of information and network. With high development of the computer and network technology, the concealment and safety of the information appears especially important. Traditional recognition technology has not been able to meet the needs of today’s information technology, therefore, effective recognition technology has become one of the important research. In nowadays, identification technology based on biological characteristics is the latest direction, while it’s one of the most important branch of biometric technology. Face recognition technology has a broader social demand and economic value, which also have important influence for the country’s security at the same time. Putting forward effective face recognition algorithm has become one of the most popular research at present.This paper study the present situation of face recognition technology via analysis and demonstration, and put forward a emotional neural network algorithm based on previous research, which has been put in face recognition technology. Experimental results show that identify effect of this method works well. This paper is mainly composed of five parts:1. Analysing about face recognition technology background and current situation detaily, and find out the advantages and disadvantages of current existing face recognition algorithm after introducing the research both at home and abroad briefly, as well as figuring out research area.2. Aim at the specialties of BP, introduce emotional factors and reestablishes neural network structure.Finally build EMBP and provide learning algorithms in detail.3. Constructing emotion structure and deducing learning algorithm steps in detail according to characteristics of RBF neural network.4. Applying emotional neural network EMBP in face recognition. Compressing and normaling singular value feature vector which was got from face image, and rearrange them in addition to achive ones with better stability and can be distinguished easily. EMBP emotional neural network is designed and performed better than ordinary BP neural network.5. Applying emotional neural network EMRBF in face recognition. Firstly, compress the facial image using PC A transformation, and then get it in low dimension by LDA. At last, EMRBF turns up. Contrast experiment shows that this method has a certain extent of the recognition rate.