MetaGait: Learning to Learn an Omni Sample Adaptive Representation for Gait Recognition

TL;DR

MetaGait introduces a meta-learning based approach to adaptively represent gait features, significantly improving recognition accuracy by addressing covariate variations and scale issues in gait recognition tasks.

Contribution

The paper proposes MetaGait, a novel meta-learning framework that learns omni-sample adaptive representations for gait recognition, enhancing model adaptiveness across multiple scales and dimensions.

Findings

Achieves state-of-the-art accuracy on CASIA-B and OU-MVLP datasets.

Effectively captures omni-scale dependencies and temporal information.

Outperforms existing gait recognition methods.

Abstract

Gait recognition, which aims at identifying individuals by their walking patterns, has recently drawn increasing research attention. However, gait recognition still suffers from the conflicts between the limited binary visual clues of the silhouette and numerous covariates with diverse scales, which brings challenges to the model's adaptiveness. In this paper, we address this conflict by developing a novel MetaGait that learns to learn an omni sample adaptive representation. Towards this goal, MetaGait injects meta-knowledge, which could guide the model to perceive sample-specific properties, into the calibration network of the attention mechanism to improve the adaptiveness from the omni-scale, omni-dimension, and omni-process perspectives. Specifically, we leverage the meta-knowledge across the entire process, where Meta Triple Attention and Meta Temporal Pooling are presented respectively to adaptively capture omni-scale dependency from spatial/channel/temporal dimensions simultaneously and to adaptively aggregate temporal information through integrating the merits of three complementary temporal aggregation methods. Extensive experiments demonstrate the state-of-the-art performance of the proposed MetaGait. On CASIA-B, we achieve rank-1 accuracy of 98.7%, 96.0%, and 89.3% under three conditions, respectively. On OU-MVLP, we achieve rank-1 accuracy of 92.4%.