Human Pose Estimation with Spatial Contextual Information

TL;DR

This paper introduces two efficient modules, CPF and PGNN, that leverage spatial contextual information to improve human pose estimation, outperforming previous methods on benchmark datasets.

Contribution

The paper proposes simple, computationally efficient modules that incorporate spatial context into human pose estimation, enhancing accuracy over prior multi-stage approaches.

Findings

Outperforms previous methods on MPII and LSP benchmarks.

CPF effectively accumulates and guides predictions across stages.

PGNN captures spatial relationships among joints via graph learning.

Abstract

We explore the importance of spatial contextual information in human pose estimation. Most state-of-the-art pose networks are trained in a multi-stage manner and produce several auxiliary predictions for deep supervision. With this principle, we present two conceptually simple and yet computational efficient modules, namely Cascade Prediction Fusion (CPF) and Pose Graph Neural Network (PGNN), to exploit underlying contextual information. Cascade prediction fusion accumulates prediction maps from previous stages to extract informative signals. The resulting maps also function as a prior to guide prediction at following stages. To promote spatial correlation among joints, our PGNN learns a structured representation of human pose as a graph. Direct message passing between different joints is enabled and spatial relation is captured. These two modules require very limited computational complexity. Experimental results demonstrate that our method consistently outperforms previous methods on MPII and LSP benchmark.