LibreDirect3DS2v11 / README.md
Xuban's picture
Mirror Direct3D-S2 v1.1 (wushuang98/Direct3D-S2@8b04a8e, direct3d-s2-v-1-1)
5df0bc1 verified
|
Raw History Blame Contribute Delete
2.31 kB
---
license: mit
library_name: libreyolo
pipeline_tag: image-to-3d
tags:
- image-to-3d
- mesh
- direct3d-s2
- libreyolo
---
# LibreDirect3DS2v11
Direct3D-S2 v1.1, DreamTech's image-to-3D model ("Direct3D-S2: Gigascale 3D
Generation Made Easy with Spatial Sparse Attention"), repackaged for
LibreYOLO. One picture of an object in, the whole object out as a mesh,
generated as a signed distance field on sparse voxels at up to 1024^3.
LibreYOLO runs it in plain PyTorch, without torchsparse, flash-attn, Triton
or custom CUDA kernels, so the same model runs on NVIDIA GPUs, Apple silicon
(MPS) and CPU.
```python
from libreyolo import LibreGen
model = LibreGen("direct3d-s2") # downloads this repo
model("chair.png").save("chair.obj")
```
```bash
libreyolo generate source=chair.png model=direct3d-s2
```
## This is a mirror, not a new model
The Direct3D-S2 weights are DreamTech's. Nothing was retrained, converted or
re-encoded; every file under `direct3d-s2-v-1-1/` is byte-identical to
upstream and keeps upstream's layout.
- Upstream repository: [`wushuang98/Direct3D-S2`](https://huggingface.co/wushuang98/Direct3D-S2)
- Upstream revision: `8b04a8eddb7a56a0f4e89fe5f5b840c7d5610c00` (folder `direct3d-s2-v-1-1`)
- Code: [DreamTechAI/Direct3D-S2](https://github.com/DreamTechAI/Direct3D-S2)
- Paper: [arXiv:2505.17412](https://arxiv.org/abs/2505.17412)
| File | Networks |
| --- | --- |
| `direct3d-s2-v-1-1/model_dense.ckpt` | dense flow transformer and dense VAE |
| `direct3d-s2-v-1-1/model_sparse_512.ckpt` | sparse flow transformer and SDF VAE, 512^3 |
| `direct3d-s2-v-1-1/model_sparse_1024.ckpt` | sparse flow transformer and SDF VAE, 1024^3 |
| `direct3d-s2-v-1-1/model_refiner.ckpt` | 512^3 SDF refiner |
| `direct3d-s2-v-1-1/model_refiner_1024.ckpt` | 1024^3 sign refiner |
| `direct3d-s2-v-1-1/config.yaml` | network and sampler configuration |
LibreYOLO reads the DINOv2 image encoder from
[`facebook/dinov2-with-registers-large`](https://huggingface.co/facebook/dinov2-with-registers-large)
(Apache-2.0) and the BiRefNet background remover from
[`LibreYOLO/LibreTRELLISl`](https://huggingface.co/LibreYOLO/LibreTRELLISl)
(MIT). SHA-256 of every file is listed in `NOTICE`.
## License
MIT, as published by DreamTech. See [`LICENSE`](./LICENSE) and [`NOTICE`](./NOTICE).