MATLAB Implementation of Face Recognition Using Third-Order Neighborhood Method

This project focuses on developing a face recognition program using MATLAB, designed to work with the Yale University face dataset and employing the third-order neighborhood method for classification. The implementation involves studying and coding fundamental face recognition principles, applying them to real-world dataset processing. Through image analysis and processing techniques, the system can identify and distinguish different facial features to achieve accurate face recognition. Key implementation aspects include preprocessing facial images, extracting relevant features using dimensionality reduction techniques, and implementing the third-order neighborhood algorithm that compares test images against three closest neighbors in the training set using distance metrics like Euclidean or cosine distance. This project provides deeper understanding of face recognition technology and establishes a foundation for future research and practical applications by demonstrating complete workflow from data loading to classification results evaluation.