References:
►Read the full article:
► Meta’s blog post:
►Singer et al. (Meta AI), 2022, “MAKE-A-VIDEO: TEXT-TO-VIDEO GENERATION WITHOUT TEXT-VIDEO DATA”,
►Make-a-video (official page):
► Pytorch implementation:
►My Newsletter (A new AI application explained weekly to your emails!):
Join Our Discord channel, Learn AI Together:
►
#makeavideo #meta #metaai
Create Content Fast with AI: read more about Frase
Artificial intelligence makes it fast & easy to create content for your blog, social media, website, and more! Jasper is the AI Content Platform that helps you and your team break through creative blocks to create amazing, original content 10X faster.
Things are getting scary and fascinating at the same time….
Here is the (in progress) Tensorflow implementation: https://github.com/soran-ghaderi/make-a-video
How do we get access?
Where do i even begin to make one? wheres the make a video button? What?
Right after my mind was blown from text to image, text to 3d comes out.
and now this..
Thanks I understand right now you can only do 5 seconds of video? Louis are you connected with any of the developers on this new AI technology? New trails are being blazed in this new frontier of AI. I am working to figure out how I can best learn new skills and apply my old skills to work in the Artificial Intelligence industry. Any suggestions would be greatly appreciated?
Wow! This is amazing. And also bad news for stock video companies / filmers. I am worried this will cause an informaton appocolyspe where video en photos are no longer valid evidence. How to prevent or mittigate this?
Very impressive. How big is the network? parameter count.
throughout the video i was wondering how far we have come ….. text to image , images to video and now text to video
this feels so motivating 😇
Given that training for this network is scaled to the order of 100s of GPUs (likely state of the art-ish) I doubt my old notebook gtx1050 is going to be chunking out amazing results any time soon unless I just use it as a dumb terminal to the cloud. (For numerous reasons I'm not a fan of the cloud.)
As the size of these models keeps increasing do you think that compute power will eventually catch back up enough to run this kind of stuff locally or will the trend forever lead us to the hallowed computers of the cloud being guarded by the great priests of the Temples of Syrinx with everyone else accessing via a device containing less compute than a box of paperclips? (asking for a friend)
I wonder if this would work better with image to video.
And IMMEDIATELY after, phenaki video appears, showcasing much longer prompts.
cool
References:
►Read the full article: https://www.louisbouchard.ai/make-a-video/
► Meta's blog post: https://ai.facebook.com/blog/generative-ai-text-to-video/
►Singer et al. (Meta AI), 2022, "MAKE-A-VIDEO: TEXT-TO-VIDEO GENERATION WITHOUT TEXT-VIDEO DATA", https://makeavideo.studio/Make-A-Video.pdf
►Make-a-video (official page): https://makeavideo.studio/?fbclid=IwAR0tuL9Uc6kjZaMoJHCngAMUNp9bZbyhLmdOUveJ9leyyfL9awRy4seQGW4
► Pytorch implementation: https://github.com/lucidrains/make-a-video-pytorch
►My Newsletter (A new AI application explained weekly to your emails!): https://www.louisbouchard.ai/newsletter/