Efficient Human Pose Estimation with Depthwise Separable Convolution and Person Centroid Guided Joint Grouping

TL;DR

This paper introduces a new efficient approach for 2D human pose estimation using depthwise separable convolutions and a person centroid-guided grouping method, achieving competitive accuracy with low computational cost.

Contribution

It proposes a novel ResBlock with depthwise separable convolution and a bottom-up multi-person pose estimation method using person centroids as roots for joint grouping.

Findings

Achieves competitive accuracy on MPII and LSP datasets.

Reduces computational costs compared to existing methods.

Effective joint grouping via centroid-guided offsets.

Abstract

In this paper, we propose efficient and effective methods for 2D human pose estimation. A new ResBlock is proposed based on depthwise separable convolution and is utilized instead of the original one in Hourglass network. It can be further enhanced by replacing the vanilla depthwise convolution with a mixed depthwise convolution. Based on it, we propose a bottom-up multi-person pose estimation method. A rooted tree is used to represent human pose by introducing person centroid as the root which connects to all body joints directly or hierarchically. Two branches of sub-networks are used to predict the centroids, body joints and their offsets to their parent nodes. Joints are grouped by tracing along their offsets to the closest centroids. Experimental results on the MPII human dataset and the LSP dataset show that both our single-person and multi-person pose estimation methods can achieve competitive accuracies with low computational costs.