Inspiration
I wanted to build an app that help us workout anywhere and anytime.
What it does
It detect an object through camera or image file and then show the prediction of object classification.
How we built it
Using a Windows ML and inference with ONNX model trained via Azure Custom Vision.
Challenges we ran into
Earlier, I am using a model that send a live video to IOT hub , but somehow it is unable to run the model for prediction. I also tried to used React for browser image analyzer but somehow the app is crashed after it analyzes a continuous loop image. Using my final framework, I also tried a few way to integrate with openCV so I can analyze the image in real time and count the repetitions but there is a few error to run the program.
Accomplishments that we're proud of
I am able to run a few inference and try to use as many framework this time.
What we learned
I learned a lot during Hackathon and it gives me so much more outlook on the machine learning and data analysis.
What's next for Personalized Gym TrAIner
I want to published my app and improve predictions tagging.
Log in or sign up for Devpost to join the conversation.