README.md

---
license: cc-by-nc-4.0
pretty_name: TwinWorld 2026 Challenge Dataset
language:
  - en
task_categories:
  - image-to-3d
tags:
  - gaussian-splatting
  - novel-view-synthesis
  - 3d-reconstruction
  - semantic-segmentation
  - remote-sensing
  - drone-imagery
  - urban-scene
  - eccv-2026
---

# TwinWorld 2026 Dataset

Drone-captured urban building scenes for the TwinWorld 2026 (ECCV) challenge:
Gaussian Splatting-based novel-view synthesis and 3D reconstruction (geometry
and semantics). The challenge data is preprocessed from two real-world
collections:

- **TUM**, from [TUM2TWIN](https://tum2t.win/)
- **Gold Coast**, from [GT-LOD3-Benchmark](https://github.com/gdslab/GT-LOD3-Benchmark)

<table>
<tr>
<td width="50%"><img src="media/Fig_TUM.png" alt="TUM2TWIN subsets used in this challenge" width="100%"></td>
<td width="50%"><img src="media/Fig_GoldCoast.png" alt="Gold Coast local visualization" width="100%"></td>
</tr>
<tr>
<td><sub>TUM2TWIN subsets used in this challenge. <b style="color:red">Red</b> = development scenes, <b style="color:blue">blue</b> = final-testing scenes.</sub></td>
<td><sub>Local visualization of the Gold Coast data.</sub></td>
</tr>
</table>

## What's included

```
Twinworld_Datasets/
├── Data_TUM/
│   └── scene_000 .. scene_008/
│       ├── train/
│       │   ├── images/*.JPG
│       │   └── sparse/0/
│       │       ├── cameras.txt
│       │       ├── images.txt
│       │       ├── points3D.txt
│       │       └── points3D.ply        # initialization point cloud
│       ├── test/
│       │   ├── images/*.JPG            # scene_000-003 only
│       │   └── sparse/0/
│       │       ├── cameras.txt
│       │       └── images.txt          # poses only, always provided
│       └── 3d_gt/
│           └── point_cloud.ply         # scene_000-003 only, x, y, z
├── Data_Goldcoast/
│   └── scene_009 .. scene_012/         # same layout as above
│       └── 3d_gt/point_cloud.ply       # scene_009-010 only, x, y, z, classification
└── script/
    └── render_test_poses.py
```

`train/` and `test/` inside one scene share the same world coordinate system
(one joint COLMAP reconstruction per scene), so a model trained on `train/`
can be rendered directly at the poses listed in `test/sparse/0/`. Camera
intrinsics/extrinsics use the standard COLMAP text format (`cameras.txt`,
`images.txt`); `points3D` is triangulated from all images (train and test
alike) and is only provided under `train/sparse/0/`.

The ground-truth point cloud (where provided) carries `x`, `y`, `z` for TUM
scenes, plus an integer `classification` for Gold Coast scenes:

| ID | Class |
|---:|---|
| 0 | ground |
| 1 | wall |
| 2 | roof |
| 3 | window |
| 4 | other |
| 255 | ignore |

## Development scenes vs. final testing scenes

| | TUM | Gold Coast | What you get |
|---|---|---|---|
| **Development** | `scene_000`–`scene_003` | `scene_009`–`scene_010` | Everything: `train/`, `test/images` + `test/sparse/0`, and `3d_gt/point_cloud.ply`. |
| **Final testing** | `scene_004`–`scene_008` | `scene_011`–`scene_012` | `train/` and `test/sparse/0` (camera poses only). No `test/images` and no `3d_gt/point_cloud.ply` are included. |

For final-testing scenes, `test/sparse/0/images.txt` still lists the full
camera pose for every held-out frame, so you always know exactly which
viewpoints to render, even though you never receive the photos or geometry
those poses were held out from. Development scenes give you everything
locally so you can self-check before relying on the final-testing scenes.

## Training example

Using [graphdeco-inria/gaussian-splatting](https://github.com/graphdeco-inria/gaussian-splatting)
(any other method/codebase works the same way, as long as it consumes
COLMAP-format input):

```bash
git clone --recursive https://github.com/graphdeco-inria/gaussian-splatting
cd gaussian-splatting

python train.py -s <path-to>/Twinworld_Datasets/Data_Goldcoast/scene_009/train -m <output_dir>
```

Do **not** pass `--eval`: the train/test split is already fixed by this
dataset's own `train/`/`test/` folders, so `train.py` should not carve out
its own held-out views from `train/`.

## Rendering test views

`script/render_test_poses.py` renders a trained model at this dataset's test
camera poses. It targets the vanilla `gaussian-splatting` repo above, so copy
it into that repo's root (next to `train.py`) and run it from there:

```bash
python render_test_poses.py \
    --model_ply <output_dir>/point_cloud/iteration_30000/point_cloud.ply \
    --camera_pose_dir <path-to>/Twinworld_Datasets/Data_Goldcoast/scene_009/test/sparse/0 \
    --output_dir <render_dir>/scene_009
```

This writes `<render_dir>/scene_009/rgb/<frame_id>.png`, one image per pose
in `test/sparse/0/images.txt`, named after that pose's own image name. If you
used a different method/codebase, adapt the model-loading and rendering calls
to your own repo, keeping the same camera-pose parsing
(`cameras.txt`/`images.txt`) and output naming.

## Checking your results with `metrics.py`

For development scenes, where the real test photos are provided locally, you
can self-check PSNR/SSIM/LPIPS with `gaussian-splatting`'s own
[`metrics.py`](https://github.com/graphdeco-inria/gaussian-splatting/blob/main/metrics.py).
It expects renders and reference photos arranged in its own `renders/`/`gt/`
layout rather than the flat `rgb/` folder `render_test_poses.py` produces, so
a quick reorganization is needed first. See `metrics.py`'s own usage notes
for the exact layout it expects.
Final-testing scenes have no local ground truth, so scores for those only
come back after official evaluation.

## References

Wysocki, Olaf, et al. "TUM2TWIN: Introducing the large-scale multimodal urban
digital twin benchmark dataset." *ISPRS Journal of Photogrammetry and Remote
Sensing* 232 (2026): 810–830.

Kim, Han Sae, et al. "GT-LOD3: LOD3 Semantic 3D Building Reconstruction
Benchmark Dataset." *ISPRS Annals of the Photogrammetry, Remote Sensing and
Spatial Information Sciences* 11 (2026): 293–302.