RepWise

RepWise is your best trainer! Log in, create a workout, and track your exercises. Get a detailed report after each session with form analysis and tips to improve. Train smarter and better everyday!"

Comment

Inspiration

We came up with the idea of RepWise because two of our three group members are regular gym members. They realize how difficult it is for beginners in the gym to struggle with maintaining proper exercise form which could lead to injury or insufficient workouts. Personal trainers are not always affordable for everyone so we created a smart, AI-driven tool that provides real-time feedback and guidance to our users. Allowing users to train safely and effectively from anywhere.

What It Does

RepWise is a web-based fitness assistant that helps users track their form. Users can:

  • Log in and create workouts by selecting exercise names and intensity levels.
  • Use their webcam to track their movements during each exercise in real time and count their reps.
  • Receive a post-workout report highlighting areas for improvement, such as form adjustments.

How We Built It

  • Frontend: Developed using React. It allows users to customize workouts, start sessions, and view feedback reports.
  • Backend: Built with Python, managing user authentication, session handling, and storage of workout data.
  • AI-Powered Tracking: Implemented mediapipe and openCV to track and analyze the user's movements using their webcam.
  • Report Generation: llama API was used to generate user reports.

Challenges We Ran Into

  • Pose Detection Accuracy: Ensuring that the AI could accurately analyze form for various body types, and camera angles was a significant hurdle. Trying to get the AI to count the rep based on the angles between the body parts was really difficult as well.
  • Learning Curve: Our member who worked on the back end had to learn openCV and mediapipe. Our member working on the front-end had to learn javascript and React.

Accomplishments That We're Proud Of

  • Successfully integrating real-time AI-powered pose tracking into a web app.
  • Building a clean and intuitive interface for users of all fitness levels.
  • Developing an automated reporting system that provides meaningful insights to help users improve their workouts.
  • Making fitness more accessible and safer for individuals without access to personal trainers.

What We Learned

  • How to use AI and computer vision to analyze human movements.
  • Strategies for optimizing performance in real-time applications using mediapipe and openCV.
  • The challenges of integrating AI-powered functionality into a web application while maintaining a smooth user experience.
  • Our group members learned how to use React, Javascript, flask and how to combine a front end with the back end

What's Next for RepWise

  • Exercise Library Expansion: Add more exercises and movement types for a wider range of users.
  • Mobile App: Develop a mobile version of RepWise to make it even more accessible.
  • Live Feedback: Implement live audio or visual cues during workouts for immediate corrections.
  • Personalization: Use machine learning to create tailored workout plans based on user performance and goals.
Share this project:

Updates