3D Pose Based Feedback For Physical Exercises

Ziyi Zhao, Sena Kiciroglu, Hugues Vinzant, Yuan Cheng, Isinsu Katircioglu, Mathieu Salzmann, Pascal Fua; Proceedings of the Asian Conference on Computer Vision (ACCV), 2022, pp. 1316-1332


Unsupervised self-rehabilitation exercises and physical training can cause serious injuries if performed incorrectly. We introduce a learning-based framework that identifies the mistakes made by a user and proposes corrective measures for easier and safer individual training. Our framework does not rely on hard-coded, heuristic rules. Instead, it learns them from data, which facilitates its adaptation to specific user needs. To this end, we use a Graph Convolutional Network (GCN) architecture acting on the user's pose sequence to model the relationship between the the body joints trajectories. To evaluate our approach, we introduce a dataset with 3 different physical exercises. Our approach yields 90.9% mistake identification accuracy and successfully corrects 94.2% of the mistakes.

Related Material

[pdf] [supp] [arXiv] [code]
@InProceedings{Zhao_2022_ACCV, author = {Zhao, Ziyi and Kiciroglu, Sena and Vinzant, Hugues and Cheng, Yuan and Katircioglu, Isinsu and Salzmann, Mathieu and Fua, Pascal}, title = {3D Pose Based Feedback For Physical Exercises}, booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)}, month = {December}, year = {2022}, pages = {1316-1332} }