This is Face Identification project done under guidance of Dr. Anand Mishra
Project Page
Project Presentation
Facial recognition technology is widely used in security, authentication, and personalized services. This project compares three techniques—Local Binary Patterns (LBP), Histogram of Oriented Gradients (HoG), and Convolutional Neural Networks (CNN)—for classifying facial images into predefined categories. We aim to evaluate the effectiveness of these techniques in terms of accuracy and computational efficiency.
Labeled Faces in the Wild (LFW) is a renowned database of face photographs designed for studying unconstrained face recognition. Developed by researchers at the University of Massachusetts, Amherst, it consists of 13,233 images of 5,749 individuals, sourced from the web and processed using the Viola Jones face detector.
For the face identification task, we have explored three distinct feature extraction techniques—Local Binary Patterns (LBP), Histogram of Oriented Gradients (HoG), and Convolutional Neural Network (CNN) features.
After conducting comprehensive experiments, we presented our findings, highlighting the best accuracy achieved for each feature extraction technique. We observed that the ANN model on HoG features yielded the highest accuracy among all techniques tested.