Pose Fit: ML Powered Fitness Application

This study explores the efficacy of MoveNet and MediaPipe frameworks in tracking Bicep Curls, Deadifts, and Squats, with the objective of enhancing fitness monitoring and promoting exercise safety. Through a comprehensive comparative analysis, the accuracy and real-time performance of these frameworks during weightlifting exercises are evaluated. The accuracy achieved with MediaPipe is 92-96% whereas with MoveNet it’s $\mathbf{8 2 - 8 8 \%}$ depending on the stability of the camera and device features. Additionally, a sophisticated mobile application is developed using Flutter, offering users a seamless interface to interact with the pose tracking models. The application facilitates exercise selection, provides real-time feedback on form and posture, and enables progress tracking. By integrating advanced pose tracking technology with mobile integration, this research aims to improve fitness experiences and mitigate the risk of injuries. The findings provide valuable insights into optimizing pose tracking frameworks for effective exercise guidance, thus contributing to the advancement of fitness technology for enhanced user outcomes and overall wellbeing. This holistic approach underscores the importance of leveraging cutting-edge technology to foster safer and more effective fitness practices.