Work

Limb movement detection and analysis based on visual recognition of human posture

Year: 2025

Type: article

Abstract: Current motion detection and evaluation technologies face challenges such as limited scalability, imprecise feedback, and lack of personalized guidance. To address these challenges, this research integrated efficient BlazePose technology with pioneering DW_KNN* algorithm, resulting in the remarkable accuracy of 98.2% in action recognition and showcasing outstanding scalability. Furthermore, the established ACLstm time series prediction model could comprehensively analyze historical sports data and associated factors of users. In Rehab dataset, MAE(Mean Absolute Error, MAE) loss was 1.383 for motion count and 0.508 for motion time. This innovative framework delivered precise feedback and tailored guidance for physical exercise and medical rehabilitation.

Source: Discover Artificial Intelligence

DOI: 10.1007/s44163-025-00246-4

Authors: Zhiguo Xiao, Chunxiang Wang, Tianjiao Ding, Xiangfeng Shen, Xinyuan Li, Dongni Li

Institutions: Beijing Institute of Technology, Changchun University

Language: English

Topic: Human Pose and Action Recognition

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $1,590