Human MotionLess Dataset (HuMoLs)

Dataset

Abstract

The HuMoLs dataset consists of 101 videos capturing a single human subject in a motionless state. The dataset can used for motion detection in healthcare monitoring settings, or other experiments.

The dataset is designed to aid in detecting inactivity or loss of consciousness in various scenarios.

To preserve anonymity, each volunteer's face has been deepfaked. Each frame may have a slightly different face.

The dataset can be used freely, but you must cite:

OPPH: A Vision-Based Operator for Measuring Body Movements for Personal Healthcare. Longfei Chen, Subramanian Ramamoorthy, Robert B Fisher. Proc. ECCV workshop/ACVR, 2024.
Date made available20 Aug 2024
PublisherEdinburgh DataShare
Geographical coverageUNITED KINGDOM,UK

Cite this