With the use of particular facial traits, law enforcement agencies and forensic artists can create criminal sketches with the help of a software system called the Criminal Sketch Generator with Graphical User Interface (GUI). Based on user-selected parameters, the system generates realistic and detailed sketches using Generative Adversarial Networks (GANs). This solution’s main objective is to offer a simple and easy-to-use interface for creating criminal sketches that can support investigations. A variety of facial traits, including hairstyles, eyeglasses, facial hair, scars, and tattoos, are available for selection by users. The GAN model uses these attributes as input to direct it in producing a artistic portrait that satisfies the required attributes. Our technology also reduces noise and restores the image’s blind areas. As a consequence, the image quality will improve, and the findings will be more accurate. The resulting picture may be further adjusted by distinguishing facial characteristics, which will be valuable for law enforcement authorities in criminal identification by providing a more promising and accurate face image of the culprit.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

AI-Powered Face Sketching for Criminal Identification

  • S. Mahesh Kumar,
  • R. Elangovan,
  • K. R. Goutham,
  • S. Jayamani

摘要

With the use of particular facial traits, law enforcement agencies and forensic artists can create criminal sketches with the help of a software system called the Criminal Sketch Generator with Graphical User Interface (GUI). Based on user-selected parameters, the system generates realistic and detailed sketches using Generative Adversarial Networks (GANs). This solution’s main objective is to offer a simple and easy-to-use interface for creating criminal sketches that can support investigations. A variety of facial traits, including hairstyles, eyeglasses, facial hair, scars, and tattoos, are available for selection by users. The GAN model uses these attributes as input to direct it in producing a artistic portrait that satisfies the required attributes. Our technology also reduces noise and restores the image’s blind areas. As a consequence, the image quality will improve, and the findings will be more accurate. The resulting picture may be further adjusted by distinguishing facial characteristics, which will be valuable for law enforcement authorities in criminal identification by providing a more promising and accurate face image of the culprit.