Feature Requests #46
Replies: 26 comments 19 replies
-
|
Implementation of DreamAvatar for high-quality 3D avatar generation from text prompts.
|
Beta Was this translation helpful? Give feedback.
-
|
it is interesting that they use a parameterized human model. |
Beta Was this translation helpful? Give feedback.
-
|
The biggest one is: 3+ image(s) guided game avatar generator. interactive ui/drag,etc or text guide edit should be better. |
Beta Was this translation helpful? Give feedback.
-
|
Image to 3D |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of Perp-Neg-stablediffusion , this seems to solve the Janus problem |
Beta Was this translation helpful? Give feedback.
-
|
Already in our plan :) |
Beta Was this translation helpful? Give feedback.
-
|
Supported! Please follow the instructions here to try it out. |
Beta Was this translation helpful? Give feedback.
-
|
Nice work! |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of TextMesh , which utilizes pseudo image supervision. |
Beta Was this translation helpful? Give feedback.
-
|
Implemented! Please follow the instructions here to give it a try. |
Beta Was this translation helpful? Give feedback.
-
|
A galley |
Beta Was this translation helpful? Give feedback.
-
|
Just an idea, but it would be cool to have a way to preview training live like the instantNERF viewer or NERFStudio? Longer term, assuming this stuff gets good enough start spending time art directing, it would be cool to be able to watch it update, and if it starts "growing" in a direction you don't like, you can reset it to the last saved checkpoint and tune the training parameters or seed, or even spawn a bunch of evolutionary branches! |
Beta Was this translation helpful? Give feedback.
-
|
Can you please add this paper method (TextMesh). Results look really nice (though could be cherypicked) and authors claim that their model beats DreamFusion. |
Beta Was this translation helpful? Give feedback.
-
|
Implemented! Please follow the instructions here to give it a try. |
Beta Was this translation helpful? Give feedback.
-
|
Reproduce ashawkey/stable-dreamfusion#263 (My team at Stability AI might help with that. ;-) ) |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of ProlificDreamer |
Beta Was this translation helpful? Give feedback.
-
|
Done. Following the instruction to have a try! |
Beta Was this translation helpful? Give feedback.
-
Wow, so efficient! |
Beta Was this translation helpful? Give feedback.
-
|
Can you improve your model support building 3d model shape or 3d part assembly? |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of Neuralangelo: High-Fidelity Neural Surface Reconstruction (https://research.nvidia.com/labs/dir/neuralangelo/) 🥺🥺 |
Beta Was this translation helpful? Give feedback.
-
|
Hi! If you simply want to use tricks from Neuralangelo to train from multi-view images, you can try my instant-nsr-pl codebase. I'm also planning to try these tricks in threestudio and see if we can get any benefit from it:) |
Beta Was this translation helpful? Give feedback.
-
|
Wow , I've tried your repository and I feel it's very good |
Beta Was this translation helpful? Give feedback.
-
|
We used techniques from Neuralangelo in our implementation of TextMesh. Welcome to give it a try following the instructions! |
Beta Was this translation helpful? Give feedback.
-
|
HiFA: High-fidelity Text-to-3D with Advanced Diffusion Guidance |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of ATT3D: Amortized Text-To-3D Object Synthesis |
Beta Was this translation helpful? Give feedback.
-
|
On it :) |
Beta Was this translation helpful? Give feedback.
-
|
I'm waiting for the implementation of ATT3D sincerely. @bennyguo |
Beta Was this translation helpful? Give feedback.
-
|
me too |
Beta Was this translation helpful? Give feedback.
-
|
are u done yet? |
Beta Was this translation helpful? Give feedback.
-
|
Guys u r crushing it. I would love to request a feature into threestudio- and that is to have the ability to start the training from a pre existing obj file to guide the shape a bit. Like a sphere obj file could begin the training and become an apple While a cylinder might start and form a banana Or if you had a type of couch as an obj file and wanted to start from that shape to make it spaceaged or something... Or to start with a sculpture obj file thats frowning and caricaturize it and make the 3d model wildly smiling with a prompt Maybe this can be done but i would find it incredibly useful to tweak existing 3d models by prompt. Btw great work guys been following dreamfusion since its release and amazed by the progress- about to write an article just about the progress |
Beta Was this translation helpful? Give feedback.
-
|
Thank you! Actually, our Fantasia3D implementation supports initializing from a custom mesh. We provide some results based on a human mesh here and you may try other custom meshes too. Looking forward to the article! |
Beta Was this translation helpful? Give feedback.
-
|
Along similar lines for NeRF-based models (DreamFusion etc), starting from some MLP checkpoint would accelerate convergence! |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of Magic123. Most of the techniques are already implemented here so it might now be much more work. |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of DreamHuman. DreamHuman: Animatable 3D Avatars from Text |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of 3DFuse : Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation project page: https://ku-cvlab.github.io/3DFuse/
|
Beta Was this translation helpful? Give feedback.
-
|
Hello, I am one of the co-authors of 3DFuse, and we have an implementation of 3DFuse with ProlificDreamer baseline (other baselines can be easily adapted as well) with many promising results. Will this be incorporated if we push our code to Threestudio, or are there any further steps to take? Thank you! |
Beta Was this translation helpful? Give feedback.
-
|
Requesting https://arxiv.org/pdf/2308.13223.pdf - EfficientDreamer: High-Fidelity and Robust 3D Creation |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of |
Beta Was this translation helpful? Give feedback.
-
|
Huggingface or Tencent demo of ProlificDreamer, please! |
Beta Was this translation helpful? Give feedback.
-
|
https://github.com/dreamgaussian/dreamgaussian implementation? |
Beta Was this translation helpful? Give feedback.
-
|
requesting Implementation for https://github.com/gsgen3d/gsgen |
Beta Was this translation helpful? Give feedback.
-
|
can something like this be implemented |
Beta Was this translation helpful? Give feedback.
-
|
are u planning to implement SyncDreamer, please? |
Beta Was this translation helpful? Give feedback.
-
|
Implementation of RODIN Diffusion: https://3d-avatar-diffusion.microsoft.com/ |
Beta Was this translation helpful? Give feedback.
and others
-
Hello, 3D enthusiasts!
I know that many of you are expecting features/models from threestudio, but since we're currently a very small team, it could be hard to meet everyone's needs. So I set up this discussion to find the most-wanted features/models.
I'll give an example to start this discussion:
Beta Was this translation helpful? Give feedback.
All reactions