Domain Specific Information Based Learning for Facial Image Forensics

Multimedia forensics involves the use of scientific methods for analyzing the contents such as audio, video, and digital images. Digital image forensics is one of the sub fields of multimedia forensics which involves any kind of digital images such as face, fingerprint etc. Security systems are one of the most sophisticated systems to protect assets and privacy in recent years. The demand for fast, accurate identification and authentication is increasing day by day. Face recognition is one of the trending topics in multimedia forensics. Several researchers have worked on addressing challenges such as pose variation, blurriness, low resolution, illumination, facial expression, viewing angle and lighting conditions in face recognition. In forensics, facial images are captured under adverse conditions; hence, face recognition in forensics is more advanced than normal face recognition. To identify the suspects involved in crime, forensic experts follow manual comparison which is a time-consuming task. The main objective of this chapter is to discuss the importance of forensics and the various universities, laboratories, and applications established in this discipline. Domain specific methods for facial image analysis in forensics and the datasets, metrics which can be used for automating domain specific methods is discussed.