A Neural Network Study of Face Recognition

The difficult subject of automatic recognition has attracted a lot of interest lately since it has so many uses in so many different industries. Face recognition is one of those difficult problems, and as of right now, no method can offer a reliable response in every circumstance. A novel method for recognizing human faces is presented in this research. This method employs a two-dimensional discrete cosine transform (2D-DCT) to compress photos and eliminate superfluous data from face photographs utilizing an image-based approach to artificial intelligence. Based on the skin tone, the DCT derives characteristics from photos of faces. DCT coefficients are calculated to create feature vectors. To determine if the subject in the input picture is “present” or “not present” in the image database, DCT-based feature vectors are divided into groups using a self-organizing map (SOM), which uses an unsupervised learning method. By categorizing the intensity levels of grayscale images into several categories, SOM performs face recognition. An image database including 25 face pictures, five participants, and five photos with various facial expressions for each subject was used to complete the evaluation in MATLAB. This method's primary benefits are its highspeed processing capacity and minimal computing demands, both in terms of speed and memory use.