Body posture alignment feedback using xAI

Title Body posture alignment feedback using xAI
Summary Extracting the body alignment in different postures and giving feedback to reduce harm during physical exercise
Keywords Explainable AI, deep learning, body posture alighment
TimeFrame 1 nov 2023 - 31 may 2024
References Vivek Anand Thoutam, Anugrah Srivastava, Tapas Badal, Vipul Kumar Mishra, G. R. Sinha, Aditi Sakalle, Harshit Bhardwaj, Manish Raj, "Yoga Pose Estimation and Feedback Generation Using Deep Learning", Computational Intelligence and Neuroscience, vol. 2022, Article ID 4311350, 12 pages, 2022.

Cooney, Martin & Pihl, J & Larsson, H & Orand, A & Aksoy, Eren. (2019). Exercising with an "Iron Man": Design for a Robot Exercise Coach for Persons with Dementia. 10.13140/RG.2.2.14286.61765.

Chaudhari, Ajay, et al. "Yog-guru: Real-time yoga pose correction system using deep learning methods." 2021 International Conference on Communication information and Computing Technology (ICCICT). IEEE, 2021.

Prerequisites Deep learning courses and interest in explainable AI (xAI)
Supervisor Cristofer Englund, Fernando Alonso-Fernandez
Level Master
Status Open

Physical exercise is an important part of a healthy lifestyle. To avoid injury, proper body alignment is important while exercising. When the injury has happened and a person is doing rehab, a physiotherapist is not always present, hence a camera-based tool that can guide a person to perform safe exercises can be a very efficient tool to recover from an injury.

The project should investigate body posture and alignment tools that use camera and lidar (e.g. ToF - time-of-flight-camera). Based on the alignment, explore methods to give the user instant feedback on how to adjust the body to obtain proper alignment to minimize harm.

The work could include 1. State-of-the-Art in body posture identification 2. State-of-the-Art in xAI for body parts adjustments 3. Implementation of a body part feature extractor 4. Comparing the extracted features with a perfect (ground truth) model 5. Generate the feedback