Project Description - KaidiLyu/InfiniteLoopers GitHub Wiki

AI Nutritional Search - a mobile app that helps users search for nutritional information using AI. The application will allow users to input food items and get detailed nutritional data. Additionally, the app may include a feature where users can take pictures of food, and the AI will recognize the food and provide nutritional information.

iOS / Android

We chose React Native as our base framework because it is easier to learn, being based on JavaScript, and it works efficiently on both iOS and Android platforms.

Expo is a React Native framework and will allow for us to develop the app quicker and easier. Some advantages include file-based routing and a standard library of React Native modules.

We will use Gemini as our language model due to its free availability, speed, and ease of integration with React/React Native. Additionally, Gemini allows for free fine-tuning, which is ideal for customizing the model to focus on food nutrition. Custom Image Recognition Model: For the image recognition feature, we plan to use large publicly available food picture datasets with human-verified labels to train our own model. This model will be used to identify food items in pictures and provide corresponding nutritional data. We are still researching the costs and technical requirements for training an accurate model with large datasets.

We will use Firebase as our backend as it easily integrates with React Native, is free to use, and has more than enough capabilities for our program.

Currently not confirmed but Expo Go will be most likely be used for the deployment of this project as we are only concerned with a few people having access to the deployment.

Group members can use whatever they prefer.

We can use Figma to draft the app’s interface to ensure that the app has an intuitive and easy-to-use user experience. And Figma provides a free version, which is suitable for small teams.

Startup