Interested in flow matching? Check out these great talks.
@lipmanya: https://t.co/2LnTmm2JL5
@msalbergo: https://t.co/8FQqA9pBJd
@AlexanderTong7: https://t.co/UZsVdPbt4N
๐The ๐ ๐๐ง๐๐ซ๐๐ญ๐ข๐ฏ๐ ๐๐ซ๐๐ฆ๐ ๐ข๐ง๐ญ๐๐ซ๐ฉ๐จ๐ฅ๐๐ญ๐ข๐จ๐ง and ๐ฅ๐จ๐จ๐ฉ๐ข๐ง๐ video generation model weights (320x512) and gradio demo of our ๐ซ๐ฎ๐๐๐๐๐ช๐ง๐๐๐๐๐ have been released!
Code - https://t.co/pRw7zy6sry
Page - https://t.co/F55LCmliep
Wow! Worth watching this. China's largest video platform #Tencentvideo (97M paying China subscribers) will begin inserting extra ads into movies/series that didn't exist in the original. #computervision
๐๐๐Our proposed #FreeNoise is accepted by #ICLR 2024!!!
It is a kind of free lunch that can generate longer videos with negligible extra time cost. It now supports three popular video generation models, VideoCrafter, AnimateDiff, and LaVie.
Project: https://t.co/5o0HBBs2pY
We're excited to introduce the upgraded VideoCrafter2, featuring major enhancements compared to VideoCrafter1:
** Simplified training: no need for numerous high-quality videos
** Better visual quality
** Better motion
** Better concept combination
Page: https://t.co/VfKCYk8TtF
Introducing ๐๐๐ป๐ฎ๐บ๐ถ๐๐ฟ๐ฎ๐ณ๐๐ฒ๐ฟ! ๐คDynamiCrafter can generate high-dynamic videos by animating open-domain still images using text prompt as addtional dynamic control.
๐Paper: https://t.co/hbKdOBmFuK
๐ Project: https://t.co/F55LCmliep