Evaluation of Deep Learning Methods and Face Detection Framework with Convolutional Neural Network

In the current technological evaluations artificial intelligence frameworks are increasing every day. Face detection has become one of the interesting areas of research where accurate face detection is required. The most commonly used face detection framework is Viola Jones algorithm using Haar feature cascade algorithm. Feature extraction is important to accurately detect the face. In order to rectify complications in feature extraction process a convolutional neural network CNN based novel robust framework is developed here. The Model considers the face images and performs three different face feature extraction method comparatively. Feature extraction using hog classifier, Face detection using Haar cascade and convolutional neural network (CNN) based feature extraction process are comparatively analyzed here. The proposed model is compared with various state of art approaches in terms of accuracy, and statistical parameters such as False rejection rate, false acceptance rate and error rate. Real time face detection has significant challenges over static face detection frameworks. The face detection models with frontal face, unique attribute detection are deeply correlated with deep convolutional neural network in the proposed model.