View-invariant gait recognition method by three-dimensional convolutional neural network
Authors:
- 23 January 2018
- journal article
- research article
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 27 (01) , 013010-11
Abstract
Gait as an important biometric feature can identify a human at a long distance. View change is one of the most challenging factors for gait recognition. To address the cross view issues in gait recognition, we propose a view-invariant gait recognition method by three-dimensional (3-D) convolutional neural network. First, 3-D convolutional neural network (3DCNN) is introduced to learn view-invariant feature, which can capture the spatial information and temporal information simultaneously on normalized silhouette sequences. Second, a network training method based on cross-domain transfer learning is proposed to solve the problem of the limited gait training samples. We choose the C3D as the basic model, which is pretrained on the Sports-1M and then fine-tune C3D model to adapt gait recognition. In the recognition stage, we use the fine-tuned model to extract gait features and use Euclidean distance to measure the similarity of gait sequences. Sufficient experiments are carried out on the CASIA-B dataset and the experimental results demonstrate that our method outperforms many other methods.Keywords
Funding Information
- China Postdoctoral Science Foundation
- National Natural Science Foundation of China (61601021)
- National Natural Science Foundation of China (61601021)
- National Natural Science Foundation of China (61370128)
- Natural Science Foundation of Beijing Municipality (L172022)
- United Foundation of Beijing Natural Science and Haidian Original Innovation (L172022)
- Program for New Century Excellent Talents in University (NCET-13-0659)
- Program for New Century Excellent Talents in University (NCET-13-0659)
- Fundamental Research Funds for the Central Universities (2016RC015)
- Fundamental Research Funds for the Central Universities (2017JBM078)
This publication has 32 references indexed in Scilit:
- Learning Representative Deep Features for Image Set AnalysisIEEE Transactions on Multimedia, 2015
- Deep learningNature, 2015
- Large-Scale Video Classification with Convolutional Neural Networks2014
- Pairwise Shape configuration-based PSA for gait recognition under small viewing angle change2011
- Towards view-invariant gait modeling: Computing view-normalized body part trajectoriesPattern Recognition, 2009
- View-independent human motion classification using image-based reconstructionImage and Vision Computing, 2009
- 3D Gait Recognition Using Multiple Cameras2006
- Gait Recognition Using a View Transformation Model in the Frequency Domain2006
- Towards a view invariant gait recognition algorithm2004
- Automatic gait recognition based on statistical shape analysisIEEE Transactions on Image Processing, 2003