I built a Dash app using Tensorflow Lite and Raspberry Pi to help with my home workouts training!
HIIT PI uses machine learning (specifically pose estimation) on edge devices to help track your HIIT workout progress in real time (~30fps). The backend runs everything locally on a Raspberry Pi while you interact with the app wherever there is a web browser connecting to the same local network as the Pi does.
The only improvement I can think of: Would it be possible to make the pose estimation software automatically recognize what kind of exercise you do at the given moment (Toe Tops, Jumping Jacks, etc.)? If there wasn’t the need to manually switch in between them, I could actually imagine using this on daily basis for my own training.
Thanks a bunch for your appreciation! That is definitely a solid direction going forward, and clearly involves deep learning intensive techniques for video understandings in my opinion. I’m not quite sure how to actually implement it yet but I’ll give it a go.