Abstract:Gait-based pedestrian identification has important applications in intelligent surveillance. From anatomical viewpoint, the physical uniqueness of human gait is physiological discriminative of individuals. Therefore, in theory, like fingerprint and face, gait is used as a biometric for pedestrian identification. However, gait-based pedestrian identification still faces multiple challenges due to a vast diversity of walking conditions and complex acquisition environment during data collection. In this paper, through combining the nonlinear dimensionality reduction by using gait manifold and the temporal feature of gated recurrent unit (GRU) together, we propose a novel gait-based pedestrian identification framework. Firstly, we design a temporal enhancement module to construct a series of frame-by-frame gait trend energy images (ff-GTEIs), which represents spatiotemporal gait characteristics and does not reduce the number of samples. Secondly, a supervised locally linear embedding (LLE) dimensionality reduction scheme is proposed, which generates a low-dimensional gait manifold for each pedestrian and transforms all ff-GTEIs into corresponding gait manifold space. Thirdly, a new pedestrian identification network based on residual GRU is proposed, which is able to identify a person by comprehensively considering the similarity between its gait and corresponding gait manifold. Finally, a series of comparative experiments are carried out based on well-known gait datasets, the experimental results show that the proposed framework for pedestrian recognition in this paper exceeds most existing methods, and has achieved an average correct recognition rate 97.4% and 99.6% based on the open-accessed gait dataset CASIA Dataset B and OU-ISIR LP dataset, respectively.

A recognition algorithm based on gait touching feature

Non-local Gait Feature Extraction and Human Identification

Gait Recognition Based on the Plantar Pressure Data of Ice and Snow Athletes

Accelerometer-based Gait Recognition Via Voting by Signature Points

Human Identification Based on Gait Manifold

Translation-Invariant Representation for Cumulative Foot Pressure Images

Gait recognition via deep learning of the center-of-pressure trajectory

Proportion-based Fuzzy Gait Phase Detection Using the Smart Insole

Human Gait Recognition Based on Self-Adaptive Hidden Markov Model

A Foot-Wearable Interface for Locomotion Mode Recognition Based on Discrete Contact Force Distribution

Gait recognition via independent component analysis based on support vector machine and neural network

Research on Abnormal Gait Recognition Algorithms for Stroke Patients Based on Array Pressure Sensing System

A Novel Algorithm of Gait Recognition

Spatio-temporal Gait Feature with Global Distance Alignment

Human Gait Recognition by Integrating Motion Feature and Shape Feature

Gait recognition using wavelet descriptors and independent component analysis

Learning Efficient Spatial-Temporal Gait Features with Deep Learning for Human Identification.

Footstep pressure signal analysis for human identification

Research on Feature Extraction Algorithm for Plantar Pressure Image and Gait Analysis in Stroke Patients

Recognizing People by Lower Limb Joint Angles During Walking

Robust Gait Recognition by Integrating Inertial and RGBD Sensors