AI Virtual Gym Trainer: Using deep Learning for Real-Time Exercise Guidance and Form Correction
The growing need for fitness services and the desire for security have resulted in the use of artificial intelligence within the fitness industry. It incorporates a virtual trainer powered by AI that offers instant guidance and feedback while exercising through a hybrid LSTM-CNN model and MediaPipe pose estimation. This model Have been trained on a fitness video dataset, the model accurately tracks body movements during the bench press and deadlifts and provides correct recommendations with a rate of over 90% accuracy in form classification. Developed in a mobile application on the Unity platform, the system provides cost-free coach-free workout guidance and improves workout security, accessibility, and effectiveness of the workout.
Authors
- Zeinab Abd El Haliem Taha
- John Medhat
Institutions
- October University of Modern Sciences and Arts (EG)
Publication Details
- Published
- 2025-07-12
- DOI
- https://doi.org/10.1109/imsa65733.2025.11167265
- Primary Topic
- Human Motion and Animation
- Type
- article
- Field-Weighted Citation Impact
- 0.00