Human face recognition using neural networks
M. H. Ahmad Fadzil, H. Abu Bakar · 2002
The paper describes the development of a human face recognition system (HFRS) using multilayer perceptron artificial neural networks (MLP). The MLP network is trained with a set of face images until it is in a "learned" state. The network is capable of classifying the face input into its class. In the case of the subject face is not one of those trained, the network will register it as unknown. The system, which takes the face image input from video camera, is also developed to detect the presence of an object in front of the camera and to search for the human facial area automatically. The detected facial area is then used as the inputs to the neural network to perform recognition.>