Robust Analytics for Video-Based Gait Biometrics

TL;DR

This paper presents robust methods for gait biometrics, including gender identification, recognition accuracy improvements, and multi-person authentication, demonstrating superior performance over existing techniques.

Contribution

It introduces novel gait analysis techniques such as Posed-Based Voting, Genetic Template Segmentation, and Bayesian Thresholding, advancing the state-of-the-art in unobtrusive biometric identification.

Findings

Outperforms existing gait recognition methods

Effective gender classification from gait data

Enhanced multi-person authentication accuracy

Abstract

Gait analysis is the study of the systematic methods that assess and quantify animal locomotion. Gait finds a unique importance among the many state-of-the-art biometric systems since it does not require the subject's cooperation to the extent required by other modalities. Hence by nature, it is an unobtrusive biometric. This thesis discusses both hard and soft biometric characteristics of gait. It shows how to identify gender based on gait alone through the Posed-Based Voting scheme. It then describes improving gait recognition accuracy using Genetic Template Segmentation. Members of a wide population can be authenticated using Multiperson Signature Mapping. Finally, the mapping can be improved in a smaller population using Bayesian Thresholding. All methods proposed in this thesis have outperformed their existing state of the art with adequate experimentation and results.