IMU2SKE: Hierarchical Contrastive Learning-Empowered Video-Based Detection of Freezing of Gait in Parkinson's Disease

Zhihua Tian
Zhihua Tian
,
Yanwei Zhao
,
Jianda Han
,
Weiguang Huo
Abstract
A hierarchical contrastive learning framework for video-based detection of freezing of gait in Parkinson’s disease.
Type
Publication
In IEEE CBS
publications

IMU2SKE introduces a hierarchical contrastive strategy for robust detection of freezing of gait from video data.

Parkinson's Disease Freezing of Gait Contrastive Learning
Zhihua Tian
Authors
Zhihua Tian
M.S. Candidate in Operations Research and Cybernetics
Zhihua Tian is an M.Sc. candidate in Operations Research and Cybernetics at Nankai University. His work focuses on AI for healthcare, computer vision, and large language models, with emphasis on Parkinson’s disease assessment from smartphone videos and motion data.