Inspiration
As a cooking enthusiast, I always find myself with extra ingredients in my pantry that I don't know what to do with. That's when I thought, wouldn't it be great if there was an AI-powered recipe finder that could suggest dishes based on the ingredients I have on hand? That's how I came up with the idea for this project.
Using a combination of machine learning techniques and web development skills, I was able to train a model that can analyze a list of ingredients and suggest possible recipes that can be cooked with them. I then built a web application using Python, Flask, and CSS to make the recipe finder accessible to anyone with an internet connection.
My hope is that this tool will not only help people reduce food waste by using up extra ingredients, but also inspire them to try new dishes and become more creative in the kitchen.
What it does
This project aims to simplify meal planning and reduce food waste by providing users with recipe suggestions based on the ingredients they have on hand. With the help of a trained AI model, the application is able to analyze a list of ingredients provided by the user and match them with recipes that use those ingredients. By providing users with creative and practical recipe ideas, this application encourages individuals to make the most of the ingredients they already have, rather than letting them go to waste or constantly buying new ones. The result is a more sustainable and cost-effective approach to meal planning, while also promoting healthy and diverse eating habits.
How we built it
Firstly, we researched methods on how to extract information from a dataset and found TF-IDF is fast and can give accurate results. Then we designed API's to make calls to the server to query and load the data to the user. Next we created a beautiful frontend using HTML5 and CSS and integrated. Then we also thought about scalability and decided to containerize the application.
Challenges we ran into
Debugging the frontend issues since no proper error message used to be shown and since our laptops dont support GPU, running advanced AI language models was not possible
Accomplishments that we're proud of
Completing our initial prototype to our satisfaction
What we learned
We learnt how to design a modular code and work on integration of backend and frontend and different models that can be used to extract information from a dataset
What's next for Recipe Finder
Extend it to use dynamically generated images for the recipe using our own AI models and have more features such as to include dietary restrictions.
Log in or sign up for Devpost to join the conversation.