Eye NutriFit

• 

  Logo

Inspiration

Most people are unaware of nutritional information and lazy to track food. Nutrition misinformation is around and fitness misconception is on the rise. Most people do not know what sucralose is in diet coke? or whats the vitamin profil in Apple's. Time to go beyond JUST calories.

What it does

Our app is focused on capturing your meals, whole food or packaged food's and extracting all the needed information in one go with complete nutritional information for everyone. With the power of AI and in collab with FDA database w were able to give the user a full in depth look at the lifestyle choices (food).

How we built it

We used HTML/CSS to build the front end. Image capturing through java script Object detection and label reading (experimented with serveral technologies but chose YOLO V8n due to time constraints) API Calling using Flask to FDA database outputting nutritional information in webpage.

Challenges we ran into

• Core tech issues: not able to successfully implement large multimodal models for label reading (openflamingo, Clip, LLaVa, InstructBlip-all recent models low dev support) -API and model latencies. -API Request challanges : databsae could not be filtered according to our requirment.
• Synchornized calling of API ## Accomplishments that we're proud of -We were able to use use latest computer vision technologies (YOLOV8) for multi object identification. -put together a working pilot with rest api calling ## What we learned
• Willing to work through and expertiment with emerging technolgies like LMM's

What's next for Eye NutriFit

-Expanding the capturing capabilties of our product which include label raeding, cooked food identification, AI analytics and tracking/dashboarding food.

Built With

• api
• flask
• json
• python
• rest