Abstract:
Human gait recognition is a behavioral biometrics method that aims to determine the identity of individuals through the manner and style of their distinctive walk. It is still a very challenging problem because natural human gait is affected by many covariate conditions such as changes in the clothing, variations in viewing angle, and changes in carrying condition. Although existing gait recognition methods perform well under a controlled environment where the gait is in normal condition with no covariate factors, the performance drastically decreases in practical conditions where it is susceptible to many covariate factors. In the first section of this dissertation, we analyze the most important features of gait under the carrying and clothing conditions. We find that the intra-class variations of the features that remain static during the gait cycle affect the recognition accuracy adversely. Thus, we introduce an effective and robust feature selection method based on the Gait Energy Image. The new gait representation is less sensitive to these covariate factors. We also propose an augmentation technique to overcome some of the problems associated with the intra-class gait fluctuations, as well as if the amount of the training data is relatively small. Finally, we use dictionary learning with sparse coding and Linear Discriminant Analysis (LDA) to seek the best discriminative data representation before feeding it to the Nearest Centroid classifier. When our method is applied on the large CASIA-B and OU-ISIR-B gait data sets, we are able to outperform existing gait methods. In addition, we propose a different method using deep learning to cope with a large number of covariate factors. We solve various gait recognition problems that assume the training data consist of diverse covariate conditions. Recently, machine learning based techniques have produced promising results for challenging classification problems. Since a deep convolutional neural network (CNN) is one of the most advanced machine learning techniques with the ability to approximate complex non-linear functions, we develop a specialized deep CNN architecture for gait recognition. The proposed architecture is less sensitive to several cases of the common variations and occlusions that affect and degrade gait recognition performance. It can also handle relatively small data sets without using any augmentation or fine-tuning techniques. Our specialized deep CNN model outperforms the existing gait recognition techniques when tested on the CASIA-B large gait dataset.
