To address the deficiencies in the field of rehabilitation movement detection

such as insufficiently comprehensive modeling of inter-joint dependencies

lack of hierarchy and multi-scale in temporal modeling

and inadequate handling of jitter caused by noise

a 3D rehabilitation movement detection method based on hierarchical multi-scale spatio-temporal-frequency collaborative Transformer is proposed.Firstly

the framework of rehabilitation movement detection is introduced.Then

the pose embedding module

spatial movement perception module

adaptive frequency encoding module

temporal trajectory modeling module

weighted feature fusion module

prediction layer

and loss function in the framework are introduced respectively.Finally

comparative experiments and ablation experiments are conducted on two datasets to verify the rationality and superiority of the proposed method.The proposed method achieves optimal performance on both datasets.The results indicate that the proposed method can comprehensively model inter-joint dependencies

consider temporal modeling with hierarchy and multi-scale

and handle the jitter caused by noise reasonably

demonstrating certain superiority.