Pinned
MBZ
1,368 posts
MBZ
@babaeizadeh
Mountain View, CA
Joined June 2009
#Veo3 further blurs the lines between reality and imagination with audio, stronger text adherence, and richer visual details.
00:00
00:00
00:00- Is predicting future rewards sufficient for achieving success in visual model-based reinforcement learning? We experimentally demonstrate that this is usually *not* the case in the online settings and the key is to predict future images too. 1/5
GIF - Replying to @babaeizadehmany people asked for Anime. but this is a potato thread. so here we go "anime style footage of two potatoes having a sword fight. cinematic, fastpaced with a lot of shotcuts"
00:00 - It's hard to promote your work when your alma mater is under siege, but here we go. Introducing Phenaki, a model that can generate minutes of videos given a story. Hopefully, it will be used for some good somewhere.1/ Today we are excited to introduce Phenaki: phenaki.github.io, short-link-to-paper, a model for generating videos from text, with prompts that can change over time, and that is able to generate videos that can be as long as multiple minutes!
GIF - Replying to @babaeizadeh"Cinematic fast paced shot of a muscle sport car drifting around a corner. It's evening. The headlights of the car is cutting through the heavy fog. The license plate is "Veo". The car is moving so fast the the pedestrians are blurry. The driver is a potato."
00:00 - Blog post on our latest experiments on visual model based reinforcement learning arxiv.org/abs/2012.04603 One of the most stable and flexible libraries that I ever worked on github.com/google-researc… with @msaffar3 @danijarh @harinidkannan @chelseabfinn @svlevine @doomieIntroducing the World Models Library, an open-source, platform-agnostic suite of tasks and tools for examination of world model design and performance in visual model-based reinforcement learning. Learn more and grab the code at goo.gle/36EXY1t
GIF - Replying to @babaeizadeh"a high energy music video. the singer is a potato and the dancers are other vegetables."
00:00 - Replying to @babaeizadeh"a documentary about the famous potato shaped casino being built in Las Vegas blvd"
00:00 - Introducing FitVid, a variational video prediction model, which is capable of severe overfitting on the common video prediction benchmarks -- while having similar parameter count as the current sota models. with @msaffar3 @SurajNair_1 @svlevine @chelseabfinn @doomie
GIF - Replying to @babaeizadeh"a training montage of a potato training hard for Olympics."
00:00
