Human Gait Datasets: A Review

Abstract

Human gait recognition is a non-intrusive distance-based second generation biometric. Human motion capture is utilized in a variety of fields to evaluate, interpret, and duplicate the wide range of movements required in daily living. To consider a new recognition system, several researchers had focused on this issue. One of the most significant advantages of this recognition above others is that it does not necessitate the attention and participation of the observed person. Many human gait datasets have been developed in the previous years, Chinese Academy of Sciences (CASIA) Gait Dataset, TUM-GAID, and Southampton University (SOTON) Gait Dataset are some of the most extensively used databases.This research paper conducts a thorough study of the publicly available gait databases for gait recognition.