Gait Recognition and Its Applications to Forensics

Abstract

Gait has been considered as a biometric cue for identifying a person, and hence, gait recognition has recently become an active research area. Unlike other biometrics, gait recognition is still applicable even if the target person is captured at a distance from the camera without being aware of the camera, and therefore, it is expected to applied to be forensics using CCTV footage. In this article, we introduce basic approaches to gait recognition, as well as recently deep-learning frameworks for it. Moreover, we address our practice of gait recognition in forensics. Finally, we conclude this article with perspectives on gait recognition in the future.