FitFormTracker | Devpost

Inspiration

Our inspiration came from the powerful capabilities of Blazepose with TensorFlow for tracking body movements. We were fascinated by the potential applications of this technology and wanted to explore how we could leverage it to make a positive impact on people's lives. We also saw the potential of integrating this with the power of GPT-4 to provide insightful feedback to users.

What it does

FitFormTracker is a web application that uses a webcam to help people exercise in a safer, better, and more enjoyable way. The application uses Blazepose and TensorFlow to track the user's movements during their workout. The tracked data is then processed and sent to GPT-4, which analyzes the data and provides feedback on how the user can improve their form and technique. This feedback is then displayed to the user in real-time, allowing them to make immediate adjustments to their workout.

How we built it

We built FitFormTracker using a combination of HTML, CSS, JavaScript, and Flask. The front-end of the application was built using HTML, CSS, and JavaScript to create a user-friendly interface that displays the webcam feed and the feedback from GPT-4. The back-end was built using Flask, which handles the communication between the front-end, the Blazepose and TensorFlow models, and GPT-4.

Challenges we ran into

One of the main challenges we faced was learning how to use Flask. This was the first time any of us had used this framework, and there was a steep learning curve. We had to spend a lot of time learning the basics and understanding how to use it effectively in our project. Additionally, integrating the different components of the project (Blazepose, TensorFlow, GPT-4, and the front-end) was also a significant challenge.

Accomplishments that we're proud of

We are proud of the fact that we were able to successfully build a working application despite the challenges we faced. We managed to integrate several complex technologies and create a product that can potentially help people improve their exercise form and technique.

What we learned

We learned a lot about Flask and how to use it to build web applications. We also gained a deeper understanding of how to work with Blazepose, TensorFlow, and GPT-4, and how to integrate these technologies to create a cohesive application.

What's next for FitFormTracker

We see a lot of potential for FitFormTracker. In the future, we would like to add more features, such as support for other exercise moves. We also see the potential for teaching martial arts or other physical activities. We believe that with more time and resources, we can make FitFormTracker an even more powerful tool for fitness and health.