Definition of Manufacturing Worker Musculoskeletal Dataset and Training and Validation of Dite-HRnet

Manufacturing businesses must conduct surveys every three years based on the guidelines for investigating harmful factors of musculoskeletal work. They have experienced many difficulties in carrying out. Therefore, there is a need for a platform for deep learning-based worker posture modelling and learning to support ergonomic musculoskeletal risk factor analysis and assessment. In this paper, Artificial Intelligence learning data for ergonomic evaluation was constructed and verified through an artificial intelligence model to solve this problem.