Key points are not available for this paper at this time.
This paper explores the advanced application of Artificial Intelligence (AI) in criminal identification through facial recognition, specifically by transforming forensic sketches into realistic photos using Deep Convolutional Generative Adversarial Networks (DCGAN). When a witness provides a description of a criminal, an expert creates a forensic sketch based on this description. By using DCGAN, this sketch is fed to into a neural network, which, after training, generates accurate, realistic facial images of the suspect. This technique significantly aids crime investigations by quickly producing detailed, high-resolution images from basic sketches, even those that are incomplete or depict various poses. The method is valuable in forensics, law enforcement, facial recognition, and security systems, enhancing the efficiency and accuracy of criminal identification
Waghmode et al. (Thu,) studied this question.