GaitMM: Multi-Granularity Motion Sequence Learning for Gait Recognition
This work improves gait recognition for biometric identification by introducing a novel method to handle motion sequence redundancy, though it appears incremental as it builds on existing datasets and tasks.
The paper tackled gait recognition by addressing the issue of data redundancy from varying step frequencies and sampling rates, proposing a multi-granularity motion representation network (GaitMM) that achieved state-of-the-art performance on CASIA-B and OUMVLP datasets.
Gait recognition aims to identify individual-specific walking patterns by observing the different periodic movements of each body part. However, most existing methods treat each part equally and fail to account for the data redundancy caused by the different step frequencies and sampling rates of gait sequences. In this study, we propose a multi-granularity motion representation network (GaitMM) for gait sequence learning. In GaitMM, we design a combined full-body and fine-grained sequence learning module (FFSL) to explore part-independent spatio-temporal representations. Moreover, we utilize a frame-wise compression strategy, referred to as multi-scale motion aggregation (MSMA), to capture discriminative information in the gait sequence. Experiments on two public datasets, CASIA-B and OUMVLP, show that our approach reaches state-of-the-art performances.