ImageGPT

Inspiration

This project was inspired by my fascination with the potential of artificial intelligence to create realistic and imaginative visual content. I wanted to explore the capabilities of modern AI models in generating images from textual descriptions, a field known as text-to-image synthesis.

What it does

This is an AI image generator, users simply just enter a prompt or click on the 'surprise me' button which automatically writes a prompt for you and click on the 'Generate' button. It will then come up with 3 AI-generated images that match your prompt.

How we built it

I’ve made this using the openai API, i’ve chosen react to make the frontend of this, just because its just really easy to create interactive user interfaces with this library. and Node.js for the backend, and it just allows me to have an efficient runtime environment that executes javascript on the server and it just allows me to handle the API requests and responses nicely, and also I can keep my API Keys safe.

Challenges we ran into

I initially ran into a problem where I could not access the AI generated images but then I realized that my free credits on openai has run out so I fixed that.

Accomplishments that we're proud of

I am especially proud of the fact that we integrated the frontend and backend well.

What we learned

Learned alot about Javascript and Node.js. I usually code in python and I am very new to js libraries.

What's next for ImageGPT

I hope to have certain parameters like 'colors-pallete, 'image-size' and etc that users can input and then ImageGPT will generate images that take into consideration those parameters.

Built With

• github
• javascript
• node.js
• openai
• react
• webstorm