21 Jan
2023
21 Jan
'23
8:01 a.m.
Everybody seems to like synthetic imagery. - genmo.ai claims to provide a text-to-video service with free demo. I haven’t visited it. - an early text-to-photorealistic-scene paper called dreamfusion came out some months ago. this is like stable diffusion but generates virtual scenes and models from text rather than just images. an open source variant is at https://github.com/ashawkey/stable-dreamfusion . i tried the colab briefly but for some reason it kept getting 0 loss for me. i didn’t look for the results so i’m not sure if that’s a bug or not.