Forensic sketch-based identification is a crucial technique in criminal investigations, especially in situations where photographic or video evidence of suspects is unavailable. Traditionally, forensic sketches are manually drawn by skilled artists based on eyewitness descriptions.
However, this process is inherently subjective, time-consuming, and prone to inaccuracies due to human memory limitations and interpretation biases. These challenges significantly reduce the reliability and efficiency of sketch-based suspect identification in real-world scenarios.

With the rapid advancement of artificial intelligence, computer vision, and deep learning, automated forensic sketch generation and recognition systems have emerged as powerful alternatives. These systems aim to bridge the modality gap between sketches and photographs by learning robust feature representations that can effectively match sketches with real facial images. Techniques such as Convolutional Neural Networks, transfer learning models like ResNet and VGG16, and advanced frameworks such as Generative Adversarial Networks have significantly improved the accuracy and efficiency of sketch-to-photo recognition.