After reading a lot about AI and machine learning been used to generate images based on text prompts – I decided to learn more about the behind the scenes. Midjourney was the first project I played with and very interesting and high quality images (art) were generated. Other great project that just came up is Avatar AI that uses DreamBooth (Fine Tuning Text-to-Image Diffusion Models).
I spent some time to try to get the Dreambooth running on a Mac M1 and PC with NVidia (Cuda) card. Finally I was able to play with Invoke.ai Stable Diffusion project and generate some cool images.
Here is their web interface and some images generated locally:
OpenArt just released their Stable Diffusion Prompt Book.