SimpleDepthPose: Fast and Reliable Human Pose Estimation with RGBD-Images

TL;DR

This paper presents a fast and reliable multi-view, multi-person pose estimation algorithm using RGBD images, demonstrating strong generalization, adaptability, and open access for further research.

Contribution

It introduces a novel RGBD-based pose estimation method that improves speed, reliability, and generalization over existing approaches.

Findings

Demonstrates strong generalization to unseen datasets

Achieves fast runtime performance

Adapts to different keypoints

Abstract

In the rapidly advancing domain of computer vision, accurately estimating the poses of multiple individuals from various viewpoints remains a significant challenge, especially when reliability is a key requirement. This paper introduces a novel algorithm that excels in multi-view, multi-person pose estimation by incorporating depth information. An extensive evaluation demonstrates that the proposed algorithm not only generalizes well to unseen datasets, and shows a fast runtime performance, but also is adaptable to different keypoints. To support further research, all of the work is publicly accessible.