DOR3D-Net: Dense Ordinal Regression Network for 3D Hand Pose Estimation

TL;DR

This paper introduces DOR3D-Net, a novel dense ordinal regression approach for 3D hand pose estimation that improves accuracy by reducing noise impact through reformulating the problem into binary classification tasks with ordinal constraints.

Contribution

The paper proposes a new dense ordinal regression framework for 3D hand pose estimation, decomposing offset regression into binary classification with ordinal constraints for better accuracy.

Findings

Significant accuracy improvements over state-of-the-art methods.

Effective noise reduction in hand joint offset predictions.

Robust performance across multiple public datasets.

Abstract

Depth-based 3D hand pose estimation is an important but challenging research task in human-machine interaction community. Recently, dense regression methods have attracted increasing attention in 3D hand pose estimation task, which provide a low computational burden and high accuracy regression way by densely regressing hand joint offset maps. However, large-scale regression offset values are often affected by noise and outliers, leading to a significant drop in accuracy. To tackle this, we re-formulate 3D hand pose estimation as a dense ordinal regression problem and propose a novel Dense Ordinal Regression 3D Pose Network (DOR3D-Net). Specifically, we first decompose offset value regression into sub-tasks of binary classifications with ordinal constraints. Then, each binary classifier can predict the probability of a binary spatial relationship relative to joint, which is easier to train and yield much lower level of noise. The estimated hand joint positions are inferred by aggregating the ordinal regression results at local positions with a weighted sum. Furthermore, both joint regression loss and ordinal regression loss are used to train our DOR3D-Net in an end-to-end manner. Extensive experiments on public datasets (ICVL, MSRA, NYU and HANDS2017) show that our design provides significant improvements over SOTA methods.