K2MUSE Dataset: Multimodal Walking Data for Rehabilitation Robotics
Researchers have just launched the K2MUSE dataset, which presents a comprehensive collection of data focused on human lower-limb walking, aiming to improve rehabilitation robotics. This dataset features various types of data, including kinematic, kinetic, amplitude mode ultrasound (AUS), and surface electromyography (sEMG). It aims to address the lack of large-scale multimodal samples in gait analysis and tackles the challenges of data collection in real-world environments. By offering crucial biomechanical information from various walking activities, the dataset is designed to enhance how lower limb rehabilitation robots interact and function. You can find all the detailed results in a paper published on arXiv (2504.14602).
Key facts
- K2MUSE dataset includes kinematic, kinetic, AUS, and sEMG data.
- It addresses gaps in existing lower limb datasets for rehabilitation robotics.
- The dataset accounts for acquisition interference in real applications.
- It provides large-scale gait samples for data-driven approaches.
- The research is published on arXiv (2504.14602).
- The dataset supports development of lower limb rehabilitation robots.
- Multimodal data deepens understanding of neuromuscular alterations.
- The dataset covers task and acquisition variability.
Entities
Institutions
- arXiv