Caption AI

Inspiration - MLH Project list, AI Hackfest twitch livestream

What it does -

Convert 2d image to 3d model Converts image to caption Converts image to meme

How we built it -

Used React for front-end Used Flask for back-end We used Blip to convert image to caption We used Shap-e to convert image to 3d model Used Gemini to convert image into meme captions

Challenges we ran into -

Latency issue with large model with heavy weights Image Resolution issues Compute time

Accomplishments that we're proud of -

Look of the front-end Ability to integrate all planned ideas Aligns with our expectations going into the project

What we learned -

Gained a better technical understanding of API's the main thing AI

What's next for Caption AI -

Absolutely Nothing, it's perfect as is!

Built With

Submitted to

AI Hackfest

Created by

Worked on gemini api with pillow.
worked on flask apis and blip model.
Responsible for most of the backend

Amarnath Sharma
I was responsible for designing and building the frontend interface of our image captioning application using React and Tailwind CSS. I focused on creating a clean, responsive user experience that allows users to upload an image, generate a caption using the backend API, and edit or copy the caption with ease. I also implemented features like dark/light mode toggle, a custom file upload button, an animated loading spinner, and copy-to-clipboard functionality

Dunni Dayo-Kayode
Cammy Phantom
Ranjit Odedra