Inspiration
We read from several news articles, and found that current eating habits among children are not really good and can later lead to more health risks when they become an adult , so to address this, we try to built an educational app that can educate children about healthy foods from an early age.
What it does
NutriPix is a mobile application that educate children about nutrition via fun facts to make learning more fun and engaging. This app have a image recognition feature that can recognize several fruits and vegetables, such as apples, bananas, grapes, and broccoli.
NutriPix also comes in the form of an API that you can try locally on your own device.
How we built it
The mobile app was built using Jetpack Compose, Camera X, and TensorFlow Lite. The API and the ML model was built using FastAPI, PostgreSQL, and Mediapipe.
Challenges we ran into
This was the first time that we customize our own model using transfer learning, so to get a good result for the model was a little bit hard.
What's next for NutriPix
We plan to develop the API and mobile app more, add more functionality to the app and api, and make a better ML model.
Built With
- camerax
- fastapi
- jetpack-compose
- mediapipe
- postgresql
- tflite
Log in or sign up for Devpost to join the conversation.