mRI: Multi-modal 3D Human Pose Estimation Dataset using mmWave, RGB-D, and Inertial Sensors

TL;DR

The paper introduces mRI, a comprehensive multi-modal dataset combining mmWave, RGB-D, and inertial sensors for 3D human pose estimation, aimed at advancing home-based health monitoring applications.

Contribution

It provides a large, synchronized multi-modal dataset specifically designed for human pose estimation and action detection in home health contexts, filling a gap in existing datasets.

Findings

Each modality's strengths are analyzed through extensive experiments.

The dataset supports benchmarks for pose estimation and action detection.

Multi-modal data improves robustness in health monitoring scenarios.

Abstract

The ability to estimate 3D human body pose and movement, also known as human pose estimation (HPE), enables many applications for home-based health monitoring, such as remote rehabilitation training. Several possible solutions have emerged using sensors ranging from RGB cameras, depth sensors, millimeter-Wave (mmWave) radars, and wearable inertial sensors. Despite previous efforts on datasets and benchmarks for HPE, few dataset exploits multiple modalities and focuses on home-based health monitoring. To bridge the gap, we present mRI, a multi-modal 3D human pose estimation dataset with mmWave, RGB-D, and Inertial Sensors. Our dataset consists of over 160k synchronized frames from 20 subjects performing rehabilitation exercises and supports the benchmarks of HPE and action detection. We perform extensive experiments using our dataset and delineate the strength of each modality. We hope that the release of mRI can catalyze the research in pose estimation, multi-modal learning, and action understanding, and more importantly facilitate the applications of home-based health monitoring.