GaitHF: Enhancing Appearance-Based Gait Recognition through Height Fused Images

Author:

Zhu Jinchang1,Sun Dayang1,Cheng Yu1,Wang Hailong1,Chen Yujing1,Chen Yaowei1

Affiliation:

1. Jilin University

Abstract

Abstract

Gait recognition, an emerging field at the intersection of computer vision and biometrics, has garnered significant attention for its potential applications in surveillance, security, and healthcare. In this paper, we present a novel method that combines appearance-based gait recognition with human height data. The proposed approach aims to enhance the accuracy and robustness of gait recognition systems by incorporating complementary features derived from both gait patterns and human height. We believe that incorporating height information can offer additional discriminative power to gait recognition models, enabling them to better distinguish individuals in various scenarios. Many gait recognition convolutional neural networks using deep learning methods have made good data progress in recent years, so we also adopt this approach, e.g., deep learning methods such as (CNN) and Recurrent Neural Networks (RNN), which can automatically learn hierarchical representations of gait and height features to capture intricate patterns and relationships. Our experiments involve a comprehensive analysis using benchmark gait datasets, demonstrating the effectiveness of the proposed approach in comparison to traditional gait recognition methods. The results highlight the potential of leveraging human height information to enhance the overall performance of gait recognition systems. Our experimental data show that the results achieved by many appearance-based gait recognition models on the CASIA-B and OU-MVLP datasets progress in most conditions after using our proposed new approach, which are eye-catching in that the average accuracy improves by 1.875% and 6% on BG and CL of CASIA-B, respectively, and the average accuracy improves on the large dataset OU-MVLP is also improved by 1.35%. Overall, our work focuses on analyzing and recognizing gait images, contributing to gait recognition. The source code and datasets can be accessed at https://github.com/ReinerBRO/GaitHF.

Publisher

Research Square Platform LLC

Reference28 articles.

1. S. Sinno, B. Hu and Y. Guan, "Gait recognition: solving the key cross-view challenge," Biometric Technology Today, vol. 2020, (4), pp. 5–7, 2020.

2. A. Karambakhsh et al., "SparseVoxNet: 3-D Object Recognition With Sparsely Aggregation of 3-D Dense Blocks," in IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 1, pp. 532–546, Jan. 2024, doi: 10.1109/TNNLS.2022.3175775.

3. "Face Sketch Synthesis Using Regularized Broad Learning System,";Li P;IEEE Transaction on Neural Networks and Learning Systems,2022

4. "Efficient Body Motion Quantification and Similarity Evaluation Using 3-D Joints Skeleton Coordinates,";Aouaidjia K;IEEE Transactions on Systems, Man, and Cybernetics. Systems,2021

5. Deep gesture interaction for augmented anatomy learning,";Karambakhsh A;International Journal of Information Management,2019

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3