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

University of Wisconsin-Madison
We need to process the camera-related data due to privacy-preserving protocol, which might delay the release. We have released dataset without camera-related modalities as well as the keypoints and actions label now. The whole dataset including camera-related will be open-sourced before the conference date.

mRI is a multi-modal human pose estimation dataset focusing on rehab movements.

Visualization of mRI from different camera poses.


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 this gap, we present mRI, a multi-modal 3D human pose estimation dataset with mmWave, RGB-D, and Inertial Sensors. Our dataset consists of over 5 million 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.

Multiple Sensing Modalities


    title={m{RI}: Multi-modal 3D Human Pose Estimation Dataset using mmWave, {RGB}-D, and Inertial Sensors},
    author={Sizhe An and Yin Li and Umit Ogras},
    booktitle={Thirty-sixth Conference on Neural Information Processing Systems Datasets and Benchmarks Track},