Comparing Face Detection and Recognition Techniques

This paper implements and compares different techniques for face detection and recognition. One is find where the face is located in the images that is face detection and second is face recognition that is identifying the person. We study three techniques in this paper: Face detection using self organizing map (SOM), Face recognition by projection and nearest neighbor and Face recognition using SVM. 1 Face detection using SOM: A self organizing map (SOM) [1, 2, 3] is a neural network based model. Each node maps a input vector to a scalar output using multiplicative weights. Typical SOM structure is a grid of neurons. The self organizing map is a dimensional reduction methods that is embed high dimensional vector to a low dimensional vector. The SOM trains models such that data points which are closer in high dimension also embed closer in lower dimension. 1.1 Learning algorithm The training of the models starts from randomizing the weights of the neurons or initialized to principal components. Initializing the weights to principal components usually converges faster than random initialization [4].