| --- |
| license: other |
| license_name: stabilityai-ai-non-commercial |
| license_link: https://huggingface.co/stabilityai/stable-virtual-camera/blob/main/LICENSE |
| base_model: stabilityai/stable-virtual-camera |
| pipeline_tag: image-to-video |
| tags: |
| - novel-view-synthesis |
| - video-diffusion |
| - 3d-gaussian-splatting |
| --- |
| |
| # GeoNVS: Geometry Grounded Video Diffusion for Novel View Synthesis |
|
|
| GS-Adapter and LoRA weights for **GeoNVS**, which injects explicit 3D Gaussian |
| geometry into camera-controlled video diffusion models. |
|
|
| [Project page](https://sites.google.com/view/minjun-kang/geonvs-eccv26) | |
| [Paper](https://arxiv.org/abs/2603.14965) | |
| [Code](https://github.com/MinJunKang/GeoNVS) |
|
|
| These weights are fine-tuned from |
| [`stabilityai/stable-virtual-camera`](https://huggingface.co/stabilityai/stable-virtual-camera) |
| (SEVA) and [CameraCtrl](https://github.com/hehao13/CameraCtrl) (SVD backbone), |
| and **inherit the Stability AI Non-Commercial Research Community License** — |
| using them requires the base models and acceptance of their licenses. |
| Non-commercial research use only. |
|
|
| ## Files |
|
|
| | Path | Backbone | Description | |
| |---|---|---| |
| | `eccv_gattn/` | SEVA | Main GeoNVS model, Adaptive Fusion (`fusion_method: gattn`): `gs_adapter_weights.pth` + `pytorch_lora_weights.safetensors` | |
| | `eccv_camctrl/` | CameraCtrl (SVD) | GeoNVS on CameraCtrl: adapter + LoRA | |
| | `eccv_base/` | SEVA | Fusion ablation: Naive Fusion (`fusion_method: concat`), use with `configs/module_config/gsadapter_eccv_base.yaml` | |
|
|
| ## Usage |
|
|
| Use with the [GeoNVS code release](https://github.com/MinJunKang/GeoNVS) — |
| `tools/scripts/download_weights_ours.sh` fetches these files into |
| `pretrained_weights/`, then: |
|
|
| ```bash |
| python demo.py --data_path <benchmarkset>/dl3dv10 --num_inputs 3 \ |
| --gs_adapter_config configs/module_config/gsadapter_eccv_gattn.yaml \ |
| --gs_adapter_weight_path pretrained_weights/eccv_gattn \ |
| --lrm_model_name vggt_iv --H=384 --W=384 --version=1.0 |
| ``` |
|
|
| ## Mirrored third-party checkpoint |
|
|
| `third_party/depthsplat-gs-base-re10kdl3dv-448x768-randview2-6-f40abc4f.pth` is |
| redistributed from [DepthSplat](https://github.com/cvg/depthsplat) |
| (MIT License, © 2024 Haofei Xu). It is the checkpoint used for the paper's |
| experiments; upstream has since replaced it with a retrained file |
| (`f8ddd845`) that measurably changes results, so the original is mirrored here |
| for reproducibility. Its MIT license governs that file. |
|
|
