Add G3Splat pretrained checkpoints

README.md

@@ -1,3 +1,100 @@
 ---
 license: mit
+tags:
+  - 3d-reconstruction
+  - gaussian-splatting
+  - novel-view-synthesis
+  - depth-estimation
+  - pose-estimation
+  - computer-vision
+  - pytorch
+  - 3dgs
+  - 2dgs
+  - feed-forward
+  - self-supervised
+language:
+  - en
+pipeline_tag: image-to-3d
+library_name: pytorch
+arxiv: "2512.17547"
 ---
+
+# G³Splat: Geometrically Consistent Generalizable Gaussian Splatting
+
+<p align="center">
+  <a href="https://m80hz.github.io/g3splat"><img src="https://img.shields.io/badge/🌐_Project_Page-007ACC?style=for-the-badge" alt="Project Page"></a>
+  <a href="https://arxiv.org/abs/2512.17547"><img src="https://img.shields.io/badge/📄_Paper-B31B1B?style=for-the-badge" alt="arXiv"></a>
+  <a href="https://github.com/m80hz/g3splat"><img src="https://img.shields.io/badge/💻_Code-181717?style=for-the-badge&logo=github" alt="GitHub"></a>
+</p>
+
+<p align="center">
+  <img src="assets/teaser.png" alt="G³Splat Teaser" width="100%">
+</p>
+
+## Model Description
+
+**G³Splat** is a pose-free, self-supervised framework for generalizable Gaussian splatting that achieves state-of-the-art performance in:
+- 🎯 **Geometry Reconstruction** - Accurate depth and mesh reconstructions
+- 📐 **Relative Pose Estimation** - No camera poses required at inference
+- 🎨 **Novel View Synthesis** - High-quality image rendering from new viewpoints
+
+## Available Checkpoints
+
+| Model | Gaussian Type | Training Data | File |
+|:------|:-------------:|:-------------:|:----:|
+| G³Splat-3DGS | 3DGS | RealEstate10K | `g3splat_mast3r_3dgs_align_orient_re10k.ckpt` |
+| G³Splat-2DGS | 2DGS | RealEstate10K | `g3splat_mast3r_2dgs_align_orient_re10k.ckpt` |
+
+## Quick Start
+
+```python
+from huggingface_hub import hf_hub_download
+
+# Download 3DGS model
+ckpt_path = hf_hub_download(
+    repo_id="m80hz/g3splat",
+    filename="g3splat_mast3r_3dgs_align_orient_re10k.ckpt"
+)
+
+# Or download 2DGS model
+ckpt_path_2dgs = hf_hub_download(
+    repo_id="m80hz/g3splat",
+    filename="g3splat_mast3r_2dgs_align_orient_re10k.ckpt"
+)
+```
+
+## Usage
+
+```bash
+# Clone the repository
+git clone https://github.com/m80hz/g3splat
+cd g3splat
+
+# Run demo
+python demo.py --checkpoint pretrained_weights/g3splat_mast3r_3dgs_align_orient_re10k.ckpt
+```
+See the [GitHub repository](https://github.com/m80hz/g3splat) for full installation and usage instructions.
+
+## Training Details
+
+- **Training Data**: RealEstate10K
+- **Resolution**: 256×256
+- **Backbones**: MASt3R (ViT-Large) and VGGT
+- **Hardware**: 24× A100 GPUs (6 nodes × 4 GPUs)
+- **Training Time**: ~6 hours
+
+## Citation
+
+```bibtex
+@inproceedings{g3splat,
+  title     = {G3Splat: Geometrically Consistent Generalizable Gaussian Splatting},
+  author    = {Hosseinzadeh, Mehdi and Chng, Shin-Fang and Xu, Yi and Lucey, Simon and Reid, Ian and Garg, Ravi},
+  booktitle = {arXiv:2512.17547},
+  year      = {2025},
+  url       = {https://arxiv.org/abs/2512.17547}
+}
+```
+
+## License
+
+MIT License

g3splat_mast3r_2dgs_align_orient_re10k.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75c8f10cb2d51ca4d9565afffaaed9fe4a1a38781e9a9722db4a3a6a5b82eb66
+size 2448474977

g3splat_mast3r_3dgs_align_orient_re10k.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b856fff90b86dbef835e6301d175fb303ddc9a4bb4faefcc12a91e1f08e1663
+size 2448477025