Accelerometer-based gait recognition via deterministic learning
Document Type
Conference Proceeding
Date of Original Version
7-6-2018
Abstract
In this paper we propose a new accelerometer-based gait recognition method, which consists of two stages: a training stage and a recognition stage. In the training stage, gait features representing gait motion dynamics, including acceleration data measured in the y-axis and z-axis of the right side of pelvis and the left thigh of the human body, are derived from accelerometers. Gait dynamics underlying different gait patterns are locally accurately modeled and approximated by radial basis function (RBF) neural networks. The obtained knowledge of approximated gait dynamics is stored in constant RBF networks. In the recognition stage, a bank of dynamical estimators is constructed for all the training patterns. Prior knowledge of gait dynamics represented by the constant RBF networks is embedded in the estimators. By comparing the set of estimators with a test gait pattern to be recognized, a set of recognition errors are generated. The average Li norms of the errors are taken as the recognition measure between the dynamics of the training gait patterns and the dynamics of the test gait pattern according to the smallest error principle. Finally, experimental results on the publicly available ZJU-GaitAcc dataset of 175 subjects demonstrated that our algorithm outperformed existing methods. By using the 2-fold and leave-one-out cross-validation styles on two subsets of this dataset, the correct recognition rates are reported to be 90.9%, 86.9% and 96.2%, 92.2%, respectively.
Publication Title, e.g., Journal
Proceedings of the 30th Chinese Control and Decision Conference, CCDC 2018
Citation/Publisher Attribution
Zeng, Wei, Jianfei Chen, Chengzhi Yuan, Fenglin Liu, Qinghui Wang, and Ying Wang. "Accelerometer-based gait recognition via deterministic learning." Proceedings of the 30th Chinese Control and Decision Conference, CCDC 2018 (2018): 6280-6285. doi: 10.1109/CCDC.2018.8408232.