I am really blown away & can't wait to start animating/creating my own design. The possibilities are endless. Thanks for bringing another great video to us. Cheers.
With the horse legs synchronizing I suspect the limitations are based around video footage that it was trained on. Constraints between objects have limitations like try to move the animals vertically off the ground and see how far it goes. Or try to rotate the vehicle around completely to see the back side and see what it does when it gets ugly.
Ohmygod the making the eyes squint - do you have any idea what implications this will have for portrait photography? 🤯 Directing subjects when photographing them is not going to be as necessary now!! That's actually huge
And there's a huge risk of limited access behind a huge paywall, some dorks even wanted it in Adobe 😂 worst cloud corpo with slogan you own/buy nothing.
This will be popular in people working with retro photography of relatives, which needed for gravestones. Many if lucky only have a one photo of grandmothers (everyone have a 4 grandmothers), which usually very bad quality.
That to animate drawings and images generated by AI and humans, to even make videos. If we put everything they are developing together, at some point just thinking about it we could even make a movie.
Can anybody PLEASE PLEASE PLEASE direct me to a download of an APP that will allow me to simply load a PKL file and explore the latent space within locally, on my own GPU? WHY is this so difficult in 2024?? Stable diffusion is everywhere - in EXE format, or just unzip the app to a folder and run it. Meanwhile you have to know how to compile in Python just to compile any app for StyleGAN, and then you have to actually RUN it in Python. Can't I get a UI please? This is supremely frustrating, can anybody help me?
I'm imaging we'd use this like photoshop together with stable diffusion to create videos by layering important objects and moving them similarly to how it's done with 3d animation and using the outpainting of the environment to match the frames created from the motion of the object layer
Please not. Adobe is one of the worst corpo after their turn into cloud = you'll never own anything scheme, forever milking. It's very commercial but in Adobe case most people who really needed it will be removed from access by huge paywall. Don't forget that there's 8 billions other people outside golden million-which can afford Adobe crap (after their normal products which worked offline forever). If you would know what business really is, any businessman knows that investment in software must be once in 10 years, if you're making it constantly=you're bad businessman. Many people in CAD software industry are also complaining about cloud ponzi schemes.
Uniting the advances in Auto-GPT, in Open-Source artificial intelligences, multimodal AI, hardware improvements with three-dimensionalization and optimizations, plugins...It is very likely that in 50 years, an AGI may exist, or that an artificial intelligence will make "ethical errors" by leaving it free
I don't believe this is real, at least done in real time, until they give you a download demo so you can try it yourself. Until then, this could all be pre-rendered IMO.
I am really blown away & can't wait to start animating/creating my own design. The possibilities are endless. Thanks for bringing another great video to us. Cheers.
especially for animation is this tool incredible
Really impressive. This is the beginning of truly personalized virtual experiences in real time.
With the horse legs synchronizing I suspect the limitations are based around video footage that it was trained on. Constraints between objects have limitations like try to move the animals vertically off the ground and see how far it goes. Or try to rotate the vehicle around completely to see the back side and see what it does when it gets ugly.
Wow, editing can't be more fun 😱
Ohmygod the making the eyes squint - do you have any idea what implications this will have for portrait photography? 🤯 Directing subjects when photographing them is not going to be as necessary now!! That's actually huge
Yes, we can easily create the Futurama “Not Sure If” meme with actual photos.
No need to pay model for shooting photographies ! 😮
And there's a huge risk of limited access behind a huge paywall, some dorks even wanted it in Adobe 😂 worst cloud corpo with slogan you own/buy nothing.
Who’s putting this software out? Where can I sign up?
Is it impressive? Yes. Does it benefit humanity? Not really. Is it potentially harmful? Definitely.
This will be popular in people working with retro photography of relatives, which needed for gravestones. Many if lucky only have a one photo of grandmothers (everyone have a 4 grandmothers), which usually very bad quality.
That to animate drawings and images generated by AI and humans, to even make videos. If we put everything they are developing together, at some point just thinking about it we could even make a movie.
Can anybody PLEASE PLEASE PLEASE direct me to a download of an APP that will allow me to simply load a PKL file and explore the latent space within locally, on my own GPU? WHY is this so difficult in 2024?? Stable diffusion is everywhere - in EXE format, or just unzip the app to a folder and run it. Meanwhile you have to know how to compile in Python just to compile any app for StyleGAN, and then you have to actually RUN it in Python. Can't I get a UI please? This is supremely frustrating, can anybody help me?
This totally blew me away
Thanks for sharing! I saw the first example in linkedn and I was mind blowed!!
Imagine George Lucas having this tech back in 1998; we’d still be waiting for The Phantom Menace to be finished (possibly a good thing).
Incredible... amazing.
When can I add it to Vlad-1111
So there's no application available to the public yet?
that's right but hopefully, they will release the code next month. Will be interesting to see the capabilities.
So you just need a single 2d photo create VR? That is creepy!
I'm imaging we'd use this like photoshop together with stable diffusion to create videos by layering important objects and moving them similarly to how it's done with 3d animation and using the outpainting of the environment to match the frames created from the motion of the object layer
Does 2 minutes paper influence?
Yes, that's a great channel to follow for latest updates. Will highly recommend
Bravo .. Thank You
This is the way.
Let's call it Dragan
what it means is that with this technology video animation will make a big step forward
Amazing
Adobe has to include somehow this in Photoshop
Please not. Adobe is one of the worst corpo after their turn into cloud = you'll never own anything scheme, forever milking.
It's very commercial but in Adobe case most people who really needed it will be removed from access by huge paywall.
Don't forget that there's 8 billions other people outside golden million-which can afford Adobe crap (after their normal products which worked offline forever). If you would know what business really is, any businessman knows that investment in software must be once in 10 years, if you're making it constantly=you're bad businessman. Many people in CAD software industry are also complaining about cloud ponzi schemes.
@@fontenbleau I understand, others can maybe do it also.
Amazing. This model would be much more helpful in working scenarios.
Uniting the advances in Auto-GPT, in Open-Source artificial intelligences, multimodal AI, hardware improvements with three-dimensionalization and optimizations, plugins...It is very likely that in 50 years, an AGI may exist, or that an artificial intelligence will make "ethical errors" by leaving it free
Digital puppeteering!
I don't believe this is real, at least done in real time, until they give you a download demo so you can try it yourself. Until then, this could all be pre-rendered IMO.
😲😮😯waw
Lets wait for the code
Amazing! Adobe has their work cut out for them...
Im sorry but its time to leave social media alone!
With google involved it won't be open source. Lol
AAAAA