.png){kind=icon}
.png){kind=icon}
Return
Benchmarking raw datasets and collaboratively-evolving processed data for markerless motion capture analysis
.png){kind=icon}
OA
AI
DOI：10.1016/j.dib.2025.112044.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We present a dataset designed for benchmarking markerless motion capture methods (from videos to joint kinematics). The dataset includes both raw and processed data. Two participants performed five tasks - walking, sit-to-stand, manual material handling, handstand hold or Y-pose (depending on the participant), and a jointly performed dance sequence. Movements were captured simultaneously recorded using 10 optoelectronic cameras (Qualisys Miqus M3, 120 Hz) and 9 video cameras (Qualisys Miqus Video, 60 Hz, 1920x1088 pixels). The raw dataset provides 3D marker trajectories and video recordings. The processed dataset includes joint kinematics obtained from both marker-based motion capture and 7 different markerless methods, contributed by multiple research teams as part of a challenge organized during a national biomechanics seminar. Additionally, the open-access GitHub repository containing processed data enables researchers to contribute new markerless methods estimated and expand the dataset collaboratively. This resource aims to facilitate benchmarking and support the development of robust markerless motion analysis methods. (c) 2025 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
Keywords:
Biomechanics
Kinematics
Marker-based
Video-based analysis
Joint angles
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
1.4
Papers:
741
Citations:
1.1W
Organization
No organization information available