C034

GAIT RECOGNITION USING PRETRAINED VISION TRANSFORMERS ON SILHOUETTE SEQUENCES

MS. LYSA LIEW WAN JIN, MS. PA PA MIN

AFFILIATION
Faculty of Information Science & Technology, Multimedia University

Description of Invention

Gait recognition is quickly emerging as an innovative form of biometric technology due to its ability to function remotely and without the need for the individual to actively engage. This project research into demonstrating the application of deep learning algorithms that use silhouette representations as input. The main objective is to overcome the problems of traditional gait recognition techniques by creating an effective framework that uses the ViT architecture.