Finding your house's potential

Usage example.

Inspiration

Our idea was to use a Generative AI to generate variations of photos of rooms, and then compare their improvement using Restb.ai's API.

What it does

With this web application, users can upload a photo of a room of their house, write a prompt to generate variations, and compare four possible improvements in the house using restb.ai's score system.

How we built it

The front-end is built with plain HTML and the back-end with Python, using the Flask library. As a Generative AI, we used InstructPix2Pix (https://github.com/timothybrooks/instruct-pix2pix) and for analysing and scoring images, restb.ai API.

Challenges we ran into

All the project was a challenge in itself, since it is the first time we make a web app and use different APIs. However, our main problems, where we had more trouble and spent more time, were generating working links for images to be uploaded to, for the restb.ai API, and finding a proper Generative AI model.

Accomplishments that we're proud of

Although it is not entirely functional and needs a lot of improvement, we really like the final state of our project. We are really proud of having been able to develop a working prototype, since we had a hard time with the problems we ran into.

What we learned

We learnt a lot from various topics and technologies, such as Github, HTML, Flask, Web applications, using APIs and IA.

What's next for Finding your house's potential

Both the User Interface and the back-end should be improved. Then, we might be able to find a more adequate Generative AI, or train it ourselves. Finally, restb.ai's API offers a lot more than the simple scoring system we used for this prototype.