More info:
- https://gizmodo.com/meta-iphone-realistic-3d-avatar-facebook-vr-ar-metavers-1849060682
- https://uploadvr.com/vr-killer-app-avatar-telepresence/
Paper: https://drive.google.com/file/d/1i4NJKAggS82wqMamCJ1OHRGgViuyoY6R/view
More info:
Paper: https://drive.google.com/file/d/1i4NJKAggS82wqMamCJ1OHRGgViuyoY6R/view
Is there somewhere a free background remover with AI support as a Windows build (not an online service)?
Krita with GMIC has background removal. So does GIMP. ImageMagick can do this as well, and has integrations for everything from Ruby to Python. You have millions of options, truth be told
Hi josephhansen,
thank you very much for your answer.
But I’ve been through all of them and none of them have convinced me.
I must add that it is not for me, but for an acquaintance.
I was thinking more of something like that.
Only with a Windows Installer that automatically downloads and installs all the dependencies.
Side note microsoft outlook sent the registration email from microsoft designer to the junk folder considering it spam… seems AI is still not very good in outlook…
Hopefully, CEB Stable Diffusion will soon also have an option to “enhance” 3D renders, but locally, thanks to the user’s GPU.
Adobe continues to extend Photoshop tools powered with AI:
Apparently, Voxon has been using Blender and Cycles to drive the development of their volumetric computing tech.
This actually goes a step farther than the old “Pepper’s Ghost” displays, but there is still a long road to what is found in Star Trek.
Was half expecting to see a chess set with animated fighting pieces, or a small crouching lady in white, with a message…
InstructPix2Pix - Learning to Follow Image Editing Instructions
https://www.timothybrooks.com/instruct-pix2pix
In my opinion, that kind of development is necessary to make tools such as Stable Diffusion useful.
Read the article and it sounds very fishy in my opinion.
To me, it looks more like they take the mesh, create some kind of parametric model out of it and then instantiate those based on text prompts. I am highly suspicious this approach in generally applicable.
I agree. There are lots of “bla bla” in the article. It’s not a “AI paper” (with details about the implementation) as we now are used to read.
Looks like more a way to place meshes from a database into a 3D scene.