Occlusion-Aware Human Pose Estimation with Mixtures of Sub-Trees

TL;DR

This paper introduces an occlusion-aware human pose estimation model using mixtures of compositional sub-trees, which improves robustness to occlusions by modeling part relationships and overlaps during inference.

Contribution

It proposes a novel hierarchical model with sub-trees learned via the Chow-Liu algorithm, incorporating occlusion handling directly into the inference process.

Findings

Outperforms state-of-the-art methods on multiple datasets

Demonstrates robustness to occlusions in complex poses

Effectively models part relationships and overlaps

Abstract

In this paper, we study the problem of learning a model for human pose estimation as mixtures of compositional sub-trees in two layers of prediction. This involves estimating the pose of a sub-tree followed by identifying the relationships between sub-trees that are used to handle occlusions between different parts. The mixtures of the sub-trees are learnt utilising both geometric and appearance distances. The Chow-Liu (CL) algorithm is recursively applied to determine the inter-relations between the nodes and to build the structure of the sub-trees. These structures are used to learn the latent parameters of the sub-trees and the inference is done using a standard belief propagation technique. The proposed method handles occlusions during the inference process by identifying overlapping regions between different sub-trees and introducing a penalty term for overlapping parts. Experiments are performed on three different datasets: the Leeds Sports, Image Parse and UIUC People datasets. The results show the robustness of the proposed method to occlusions over the state-of-the-art approaches.