Blender synthetic -- Hotdog

blender_hotdog · Blender / NeRF synthetic · planned — not present locally yet

Ground-truth mesh

Not present locally. See the acquisition plan below.

Why this sample is in the benchmark

Thin cylindrical structure with sharp colour boundaries. Pairs with our local capsule sample to separate "thin geometry" effects from dataset effects. Ships exact poses, so it isolates geometry error from pose error.

Ground-truth mesh must be exported from the released .blend file.

Ground truth

GT type

exact

licence

MIT / research (see release)

source

Blender / NeRF synthetic

builder

- {'scene': 'hotdog'}

views

100 × 800px (orbit, fov 50°)

lighting

provided

Acquisition plan

kind script, automatic False, estimated size 50 MiB

1. Auto-download through nerfbaselines (recommended)

   pip install nerfbaselines && python -c "import nerfbaselines; nerfbaselines.download_dataset('blender')"

2. Original NeRF synthetic + LLFF data release
   https://drive.google.com/drive/folders/1cK3UDIJqKAAm7zyrxRYVFJ0BRMgrwhh4
3. Export ground-truth mesh from the .blend file

   blender -b hotdog.blend --python examples/export_blend_mesh.py

Blender / NeRF synthetic scenes (Mildenhall et al.): lego, chair, drums, ficus, hotdog, materials, mic, ship. 100-200 upward-facing views with exact poses (transforms.json) AND the .blend files, from which the ground-truth mesh can be exported exactly. This is the reference object-level geometry benchmark.

Comparisons

No reconstruction compared yet. Run geombench compare blender_hotdog --mock or point it at a real 3DGS output.