Depth-Based 3D Hand Pose Estimation: From Current Achievements to Future Goals

TL;DR

This paper reviews the current state of 3D hand pose estimation from depth images, analyzing top methods and identifying key challenges and future directions for improving accuracy, especially in extreme views and occlusions.

Contribution

It provides a comprehensive analysis of top 3D hand pose estimation methods, highlighting their strengths, weaknesses, and areas needing further research.

Findings

Volumetric CNNs outperform 2D CNNs in capturing spatial depth data.

Current methods achieve low errors in moderate view points but struggle with extreme angles.

Explicit modeling of structure constraints improves occlusion handling.

Abstract

In this paper, we strive to answer two questions: What is the current state of 3D hand pose estimation from depth images? And, what are the next challenges that need to be tackled? Following the successful Hands In the Million Challenge (HIM2017), we investigate the top 10 state-of-the-art methods on three tasks: single frame 3D pose estimation, 3D hand tracking, and hand pose estimation during object interaction. We analyze the performance of different CNN structures with regard to hand shape, joint visibility, view point and articulation distributions. Our findings include: (1) isolated 3D hand pose estimation achieves low mean errors (10 mm) in the view point range of [70, 120] degrees, but it is far from being solved for extreme view points; (2) 3D volumetric representations outperform 2D CNNs, better capturing the spatial structure of the depth data; (3) Discriminative methods still generalize poorly to unseen hand shapes; (4) While joint occlusions pose a challenge for most methods, explicit modeling of structure constraints can significantly narrow the gap between errors on visible and occluded joints.