Inspiration

I started playing with stable-diffusion-based video generation in November last year and fell in love with the technology but saw that existing tools were kind of complicated to use. My goal was to build a nice tool that offers high control while maintaining ease-of-use.

What it does

It is a video generation tool that looks like a classical video editor, in which you generate frames from a text prompt.

How I built it

Built it with two backends (one for DB one for GPU) and a react/nextJS frontend.

Challenges I ran into

The cloud stuff was/is really hard! Especially in times of high usage.

Accomplishments that I'm proud of

I am a physicist and started this project with almost zero frontend skills. I learnt quite a lot and am proud of the video editor timeline (although there is still lots to improve)

What I learned

ChatGPT is your friend when building something new.

What's next for neural frames

Trying to get it profitable by implementing one new feature at a time.

Built With

Share this project:

Updates