Meal Snap

Suggested Recipes List
Text & Image Input
Image Input

Inspiration

We were inspired by the common struggle of deciding what to cook with what’s already in the fridge. We wanted to make meal planning easier, reduce food waste, and save time by offering a solution that instantly suggests recipes based on available ingredients.

What it does

Our app lets users take a photo of their fridge, identifies the ingredients inside using a YOLO v8 object detection model, and suggests recipes based on those ingredients. It helps reduce food waste, save time, and cook meals without needing extra trips to the store.

How we built it

We built the app using a machine learning YOLO v8 model for object detection, combined with a large language model (LLM) to generate recipe suggestions based on the detected ingredients. The backend handles image processing, ingredient identification, and recipe suggestion logic, while the front-end is designed for an easy and intuitive user experience.

Challenges we ran into

Tuning the object detection model to accurately identify a wide range of ingredients.
Integrating the machine learning model with the natural language processing system to generate relevant and practical recipes.
Ensuring the app works seamlessly across different devices with varying camera qualities.
Fine-tuning the app to handle diverse kitchen setups and ingredient variations.

Accomplishments that we're proud of

Successfully integrating the YOLO v12 object detection model for precise ingredient identification.
Creating a user-friendly interface that’s quick and intuitive for anyone to use.
Offering an AI-powered solution that saves time and reduces food waste.

What we learned

Combination of machine learning with natural language processing for real-world applications.
Fine-tuning an object detection model for real-world environments