Advances in Dimensionality Reduction Techniques for Face Recognition: A Comparative Analysis

Face recognition technology has attracted substantial interest owing to its great potential in different applications. However, with the rapid growth in the resolution and dimensions of images represented as matrices, processing them for feature extraction has become difficult. In this paper, we examine various techniques of dimensionality reduction on image data to extract prominent features that will improve the efficiency and accuracy of recognition systems. Additional preprocessing processes were performed on the images in the dataset to improve performance. The initial section of the paper's methodology focuses on implementing several dimensionality algorithms such as PCA, 2DPCA, and 2D(Square) PCA. Subsequently, we explore facial recognition technology such as Eigenfaces. Finally, when the models' performance is evaluated using criteria such as accuracy and training time, an obvious disparity is observed.