TasteLatino

Inspiration

We wanted to allow people to understand and explore latino food culture while also educating fellow latinos on possible different dish choices with certain ingredients.

What it does

Our mobile app has multiple features, the main one being a camera that captures a picture of an ingredient and uses a machine learning model to recognize the item. It then uses an Ai chatbot to generate multiple latin recipes using said item. It also offers a recipe of the day tab which highlights a popular dish everyday for inspiration. The profile feature allows you to favorite recipes you are interested for future use. The History feature allows you to look back on items you have previously scanned.

How we built it

Our team created the app mainly on xcode using swift, including all of the front-end and back-end development . We created our own machine learning model and trained it using CoreML which we then implemented into our xcode project. We also linked a chat-bot from OpenAi into our project using an API key.

Challenges we ran into

The major challenge we stumbled upon was a fault in our original machine learning model, that was providing inaccurate image detection results. We continued to make changes to the model code but eventually decided to switch from the model we were using: MobileNetV2 to a model we created that detects and processes pictures of food ingredients.

Accomplishments that we're proud of

Major accomplishment's we are holding our head high on are the machine learning model we created to predict the ingredients the camera picture was presenting and the AI recipe creation connectivity that links the image taken to the OpenAI API by converting the image into a JSON string that reads it to the API and generates the recipe for it.