LG-Hand: Advancing 3D Hand Pose Estimation with Locally and Globally Kinematic Knowledge

TL;DR

LG-Hand is a novel 3D hand pose estimation method that integrates spatial-temporal graph neural networks with kinematic constraints, significantly improving accuracy by leveraging local and global hand structure information.

Contribution

The paper introduces Angle and Direction loss functions to incorporate local and global kinematic knowledge into 3D hand pose estimation.

Findings

Achieves promising results on FPHAB dataset

Demonstrates effectiveness of kinematic constraints in pose estimation

Shows ablation study confirms contribution of proposed losses

Abstract

3D hand pose estimation from RGB images suffers from the difficulty of obtaining the depth information. Therefore, a great deal of attention has been spent on estimating 3D hand pose from 2D hand joints. In this paper, we leverage the advantage of spatial-temporal Graph Convolutional Neural Networks and propose LG-Hand, a powerful method for 3D hand pose estimation. Our method incorporates both spatial and temporal dependencies into a single process. We argue that kinematic information plays an important role, contributing to the performance of 3D hand pose estimation. We thereby introduce two new objective functions, Angle and Direction loss, to take the hand structure into account. While Angle loss covers locally kinematic information, Direction loss handles globally kinematic one. Our LG-Hand achieves promising results on the First-Person Hand Action Benchmark (FPHAB) dataset. We also perform an ablation study to show the efficacy of the two proposed objective functions.