About the Project

The DeepAI Image Generator was created using the text-to-image model trained by DeepAI.org. The model is able to take text inputs (formatted as a string) and generates an original image based on existing images related to the input string. While some prompts create terrifylingly accurate creations, others miss the mark substantially, suggesting that the model needs further training with a wider variety of stimulus images.

This project was created in 48 hours for the Coder Academy Flex Track Hackathon, May 2021.

The Team

a picture of Theo


Theo is an artist-turned-developer who is passionate about bringing diversity to the tech world. His key areas of interest are machine learning and web accessibility. He is the proud father of a bearded dragon named Diego who is a very good debugger.

A picture of Kathryn


Kathryn a developer who doesn't know how to give a concise and interesting blurb about herself so will leave it for later. She loves tabletop games?

A picture of Mimi


Mimi is an up-and-coming developer with an avid interest in frontend and the harmony between usability and aesthetics. She is inspired by the people and world around her.