Dreamix: Video Diffusion Models are General Video Editors
https://dreamix-video-editing.github.io/

@douglass another interesting diffusion based video paper that came out recently.

Text-based video editing approach using diffusion models. Fuses low-res video info with synthesized info to match the text prompt. Method has a mixed objective with full temporal attention and attention masking to improve motion editability. Includes a framework for image animation and subject-driven video generation.

፨ Legal Entities for Web3 & DAOs 🎙️Host of Just DAO It! Podcast & /daos Channel 💼 Founder of MIDAO 🟪 Purple 🏴 bDAO nf.td/thethriller

@thethriller https://github.com/showlab/Tune-A-Video

Is anything like this already available for purchase/use?