bleedingedge.ai Profile picture
Mar 3, 2023 10 tweets 3 min read
Researchers release ControlNet: control large diffusion models using inputs beyond just text prompts.

You input:
edge maps, poses, etc + a text prompt

to output: highly specific images from models like Stable Diffusion.

Here's how people are using ControlNet 🧵: Accurately generating hands and fingers:
Jan 27, 2023 5 tweets 2 min read
Google announces MusicLM: a model to generate music from text. Here are some crazy things it can do:

1. Given audio of a melody, it can generate new music inspired by that melody customized by prompts! Here's someone humming bella ciao turned into a cappella chorus, EDM, etc. 2. Generate audio with stories and progression:
electronic song played in a videogame (0:00-0:15)
meditation song played next to a river (0:15-0:30)
fire (0:30-0:45)
fireworks (0:45-0:60)
Dec 1, 2022 11 tweets 3 min read
These are the most impressive chats we've seen with ChatGPT so far. It can... Do your homework: