IGF-Bench / README.md
igfbench-neurips2026's picture
Add dataset_card.md to top-level metadata tree in README
3a9bfc9 verified
|
Raw
History Blame Contribute Delete
21.5 kB
---
license: cc-by-nc-sa-4.0
language:
- en
pretty_name: IGF-Bench (Indoor Geometric Fidelity Benchmark)
tags:
- 3d-scene-understanding
- monocular-depth-estimation
- controlnet
- diffusion-models
- benchmark
- evaluation
- indoor-scenes
- geometric-fidelity
- 3d-front
- vanishing-points
- structural-evaluation
task_categories:
- depth-estimation
- image-to-image
- text-to-image
size_categories:
- 10K<n<100K
configs:
- config_name: default
data_files: "evaluation/igf_summary.json"
---
# IGF-Bench: Indoor Geometric Fidelity Benchmark
> **Anonymous mirror for NeurIPS 2026 Evaluations and Datasets Track double-blind review.**
> The de-anonymised author/maintainer information will replace this header at camera-ready.
**IGF-Bench** is the first benchmark for evaluating **structural-level geometric fidelity** of conditionally generated indoor scene images, going beyond perceptual metrics like FID and LPIPS. It pairs **3,600 calibrated synthetic ground-truth views** with **21,600 generated images** from six state-of-the-art ControlNet models, plus **25,200 monocular-depth estimates**, all evaluated with four complementary geometric metrics: **planarity (`L_plane`)**, **orthogonality (`L_ortho`)**, **edge alignment (`L_edge`)**, and **vanishing-point consistency (`L_vp`)**.
| Quick stats | Value |
|:---|:---|
| Calibrated GT views | 3,600 (300 rooms × 3 complexity levels × 4 viewpoints) |
| Paired generated images | 21,600 (6 ControlNet models, all conditioned on identical Canny maps) |
| Paired depth estimates | 25,200 (DepthPro on all GT + generated; DAv2 / ZoeDepth subsets) |
| Camera FOV | 90° |
| Render resolution | 1024×1024 |
| Total size | ≈ 219 GB |
| License | CC BY-NC-SA 4.0 (data) + Apache 2.0 (code) |
| Code repo | <https://anonymous.4open.science/r/IGF-Bench-Code> |
| Paper | NeurIPS 2026 E&D Track (under review) |
---
## Table of Contents
1. [What's in this Repository](#whats-in-this-repository)
2. [Headline Findings (verified against this release)](#headline-findings)
3. [Quick Start](#quick-start)
4. [Directory Structure](#directory-structure)
5. [Reproducing Each Paper Experiment](#reproducing-each-paper-experiment)
6. [Pre-Computed Result JSONs (no compute needed)](#pre-computed-result-jsons)
7. [Schema Reference](#schema-reference)
8. [License Chain](#license-chain)
9. [Citation](#citation)
---
## What's in this Repository
This HuggingFace dataset repository ships **everything needed to evaluate or
extend IGF-Bench**:
-**Ground-truth 3D-FRONT renders** with calibrated cameras (RGB + EXR depth + 7-class semantic mask + Canny edge map) at 3 complexity levels (`L0_empty` / `L1_basic` / `L2_full`).
-**Paired generated images** from six ControlNet pipelines (SD 1.5, SDXL 1.0, SD 3.5 Large, Flux.1 Dev, Hunyuan-DiT, Kolors) under a uniform protocol (Canny conditioning, seed=42, no negative prompts, `cn_scale=1.0`).
-**Per-view depth estimates** from DepthPro on all 25,200 images, plus DAv2 and ZoeDepth on the 200-view ablation subset (App C.1).
-**Pre-computed evaluation JSONs** that already populate every table and figure in the paper (`evaluation/`).
-**A trained LoRA adapter** (`experiments/finetune/dav2_lora_adapter/`) that demonstrates IGF-Bench is usable as supervision for fine-tuning a pretrained MDE (App C.8).
-**A 200-view MDE consistency subset** with DAv2 + ZoeDepth depth pre-computed for paper App C.1 (`tab:mde_ablation`).
- ✅ A formal **Datasheet for Datasets** (`DATASHEET.md`) and **Croissant 1.0 metadata** (`croissant.json`).
---
## Headline Findings
> Each finding here is reproducible from the JSONs shipped under `evaluation/` — no recompute required.
1. **Perceptual quality ≠ geometric fidelity.** Models with comparable FID (e.g., SDXL FID = 64.2 vs. Flux.1 FID = 56.3) can still differ by ≈2× in `ΔL_ortho` (Flux.1 0.032 vs. SDXL 0.061). Perceptual metrics alone miss structural failures.
2. **Conditioning architecture matters.** Flux.1's channel-concatenation conditioning achieves the lowest `ΔL_ortho` (0.032), substantially better than the residual-injection ControlNet variants used by the other five models.
3. **Edge alignment is the universal bottleneck.** All six models show 68%–89% relative `L_edge` degradation versus GT — even the best model fails on fine geometric structure.
4. **Cleanliness of synthetic GT.** DepthPro AbsRel on IGF-Bench GT is 0.056, vs. 0.084 on NYU-v2 (real Kinect) — the lower MDE-noise floor makes IGF-Bench's relative-degradation (`ΔL_*`) design more sensitive to generation-induced artefacts.
5. **`L_ortho` works as a label-free quality signal.** Picking the per-view best of 5 random Flux.1 seeds by minimum `L_ortho` cuts `L_ortho` by 35.1% relative to the best fixed seed (App C.7.1).
6. **The synthetic supervision transfers (in-domain).** Fine-tuning DepthAnything V2-Small with a 1.75%-parameter LoRA adapter improves in-domain AbsRel from 1.044 to 0.081 (−92%) on the 3D-FRONT held-out test set (App C.8).
---
## Quick Start
### Option A — Verify the paper without downloading bulk data
The full evaluation already shipped in `evaluation/` is < 50 MB. You can verify
every paper number by reading the JSONs directly:
```python
import json
from huggingface_hub import hf_hub_download
# Example: verify the GT L_plane = 0.056 number
fp = hf_hub_download(
repo_id="igfbench-neurips2026/IGF-Bench",
filename="evaluation/igf_summary.json",
repo_type="dataset",
)
data = json.load(open(fp))
gt = next(d for d in data if d["model"] == "GT_baseline")
print(round(gt["l_plane_residual_mean"], 3)) # → 0.056
```
### Option B — Download everything and re-run
```bash
pip install huggingface_hub
huggingface-cli download igfbench-neurips2026/IGF-Bench --repo-type dataset \
--local-dir ./igf-bench-data
export IGF_BENCH_ROOT=$(pwd)/igf-bench-data
```
Then clone the code at <https://anonymous.4open.science/r/IGF-Bench-Code> and
run:
```bash
python scripts/evaluate_igf.py \
--renders_root $IGF_BENCH_ROOT/renders_textured \
--generated_root $IGF_BENCH_ROOT/generated \
--depth_root $IGF_BENCH_ROOT/depth_results \
--output_summary igf_summary.json
```
This reproduces the full Table 3 from the paper. Runtime ≈ 6 hours on a single
RTX 4090; 99% of the time is spent in I/O reading per-view depth NPYs.
### Option C — Download only what you need for one row of the main table
For example, to recompute only the **Flux.1** row of Table 3 (≈ 70 GB):
```bash
huggingface-cli download igfbench-neurips2026/IGF-Bench --repo-type dataset \
--local-dir ./igf-bench-data \
--include "renders_textured/*" \
"generated/flux1_canny/*" \
"depth_results/gt/depthpro/*" \
"depth_results/gen/flux1/depthpro/*"
```
---
## Directory Structure
```
igfbench-neurips2026/IGF-Bench/
├── README.md ← this file
├── DATASHEET.md ← formal Datasheets-for-Datasets record
├── croissant.json ← MLCommons Croissant 1.0 metadata
├── dataset_card.md ← short HF dataset card (4.4 KB summary)
├── LICENSE ← CC BY-NC-SA 4.0
├── selected_rooms.json ← canonical 300-room list
├── room_statistics.json ← summary statistics
├── renders_textured/ ≈ 36 GB — Ground-truth rendered views
│ └── {scene_id}_{room_type}-{room_id}/
│ └── {L0_empty,L1_basic,L2_full}/
│ └── view_{0,1,2,3}/
│ ├── rgb_textured.png RGB render
│ ├── depth.exr metric depth, OpenEXR float32
│ ├── depth.png 8-bit depth visualisation (QA only)
│ ├── depth_gt.npy NumPy float32 cache of depth.exr (LoRA training I/O)
│ ├── canny.png Canny edges (thresholds 100/200)
│ ├── semantic_id.png 7-class IDs (0..6)
│ ├── semantic_mask.png colour-mapped semantic (QA only)
│ ├── wireframe_3d.png 3D wireframe overlay (QA only)
│ └── camera.json intrinsics + extrinsics
├── generated/ ≈ 36 GB — Generated images per model
│ │
│ │ --- Main protocol: 6 models × 3,600 views = 21,600 PNG (Table 3) ---
│ ├── sd15_canny/ SD 1.5 native 512² (300 rooms × 3 levels × 4 views)
│ ├── sdxl_canny/ SDXL 1.0
│ ├── sd35_canny/ SD 3.5 Large
│ ├── flux1_canny/ Flux.1 Dev
│ ├── hunyuan_canny/ Hunyuan-DiT
│ ├── kolors_canny/ Kolors
│ │
│ │ --- Ablation: negative prompt, App C.3 (3,552 paired views) ---
│ ├── sd15_canny_with_neg/sd15_canny/ SD 1.5 with the legacy negative prompt
│ ├── sdxl_canny_with_neg/sdxl_canny/ SDXL with the legacy negative prompt
│ │
│ │ --- Ablation: resolution, App C.4 (3,552 paired views) ---
│ ├── sd15_upsampled/sd15_canny/ SD 1.5 outputs bicubic-upsampled to 1024²
│ │
│ │ --- Ablation: seed sensitivity (App C.6) + quality gating (App C.7), 200 views each ---
│ └── ablation/
│ ├── flux1_seed123/ App C.6 / C.7 — Flux.1 with seed=123
│ ├── flux1_seed456/ App C.6 / C.7 — Flux.1 with seed=456
│ ├── flux1_seed789/ App C.7 only — Flux.1 with seed=789
│ ├── flux1_seed1024/ App C.7 only — Flux.1 with seed=1024
│ ├── sdxl_w050/ App C.7 — SDXL with cn_scale=0.50
│ ├── sdxl_w075/ App C.7 — SDXL with cn_scale=0.75
│ ├── sdxl_w125/ App C.7 — SDXL with cn_scale=1.25
│ └── sdxl_w150/ App C.7 — SDXL with cn_scale=1.50
│ (each leaf directory contains the same {room}/{level}/view_{0..3}.png structure)
├── depth_results/ ≈ 150 GB — Per-MDE depth estimates (NPY)
│ ├── gt/ Ground-truth-render depth (3 MDE backbones)
│ │ ├── depthpro/ 3,600 NPY (full set)
│ │ ├── dav2/ ablation 200-view subset (App C.1)
│ │ └── zoedepth/ ablation 200-view subset (App C.1)
│ │
│ └── gen/ Generated-image depth
│ │
│ │ --- Main protocol: 6 models × 3,600 = 21,600 NPY ---
│ ├── sd15/depthpro/ 3,600 NPY
│ ├── sdxl/depthpro/ 3,600 NPY (also dav2/, zoedepth/ for App C.1 200-view subset)
│ ├── sd35/depthpro/ 3,600 NPY
│ ├── flux1/depthpro/ 3,600 NPY (also dav2/, zoedepth/ for App C.1 200-view subset)
│ ├── hunyuan/depthpro/ 3,600 NPY
│ ├── kolors/depthpro/ 3,600 NPY
│ │
│ │ --- Ablation depth maps ---
│ ├── sd15_neg/depthpro/ 3,552 NPY (App C.3 negative prompt)
│ ├── sdxl_neg/depthpro/ 3,552 NPY (App C.3 negative prompt)
│ ├── sd15_upsampled/depthpro/ 3,552 NPY (App C.4 resolution)
│ ├── flux1_seed123/depthpro/ 200 NPY (App C.6 / C.7 seed=123)
│ ├── flux1_seed456/depthpro/ 200 NPY (App C.6 / C.7 seed=456)
│ ├── flux1_seed789/depthpro/ 200 NPY (App C.7 seed=789)
│ ├── flux1_seed1024/depthpro/ 200 NPY (App C.7 seed=1024)
│ ├── sdxl_w050/depthpro/ 200 NPY (App C.7 cn_scale=0.50)
│ ├── sdxl_w075/depthpro/ 200 NPY (App C.7 cn_scale=0.75)
│ ├── sdxl_w125/depthpro/ 200 NPY (App C.7 cn_scale=1.25)
│ └── sdxl_w150/depthpro/ 200 NPY (App C.7 cn_scale=1.50)
│ (each: {room}/{level}/{view}.npy)
├── evaluation/ ≈ 41 MB — Pre-computed metric outputs
│ ├── igf_summary.json per-model aggregated (Table 3 source)
│ ├── igf_results.json per-view detailed (≈ 25 MB)
│ ├── error_decomposition.json App C.2
│ ├── mde_ablation_*.json App C.1
│ ├── neg_prompt_ablation.json App C.3
│ ├── n1_resolution_ablation.json App C.4
│ ├── n3_lvp_improved.json App C.5
│ ├── seed_ablation.json App C.6
│ ├── wilcoxon.json per-pair Wilcoxon + Bonferroni (paper Table 3)
│ ├── anova.json two-way Type II ANOVA (paper §4.3)
│ └── fid_lpips.json perceptual baselines
└── experiments/finetune/
└── dav2_lora_adapter/ ★ 1.8 MB — Trained LoRA (App C.8)
├── adapter_config.json peft config: r=16, α=32, Q/K/V, dropout=0.05
├── adapter_model.safetensors 442,368 trainable params (1.75%)
└── README.md load instructions
```
---
## Reproducing Each Paper Experiment
| Paper Section | Required HF subsets | Expected runtime | Code entrypoint |
|:---|:---|:---:|:---|
| **§4 / Table 3** main IGF metrics | `renders_textured/`, `generated/{all}/`, `depth_results/{gt,gen}/depthpro/` | ~6 h | `scripts/evaluate_igf.py` |
| **App C.1** MDE robustness | + `depth_results/{gt,gen}/{dav2,zoedepth}/` (200 views) | ~30 min | `scripts/evaluate_mde_ablation.py` |
| **App C.2** Error decomposition | already in `evaluation/error_decomposition.json` | <1 min | `scripts/error_decomposition.py` |
| **App C.3** Negative-prompt | needs the legacy generations (archived under `generated/sd15_canny_with_neg/sd15_canny/` and `generated/ablation/sdxl_with_neg/sdxl_canny/` on HF) | ~3 h | `scripts/evaluate_neg_prompt.py` |
| **App C.4** Resolution confound | `generated/sd15_canny/` + local bicubic upsample to 1024² | ~1 h | `scripts/run_n1_resolution_ablation.py` |
| **App C.5** VP 2D | `renders_textured/` + `depth_results/{gt,gen}/depthpro/` | ~30 min | `scripts/run_n3_lvp_improved.py` |
| **App C.6** Seed sensitivity | `generated/ablation/flux1_seed{123,456}/` (already on HF) | ~30 min | `scripts/evaluate_seed_ablation.py` |
| **App C.7** Quality gating | `generated/ablation/flux1_seed{123,456,789,1024}/` (seed=42 reuses `generated/flux1_canny/`) + `generated/ablation/sdxl_w{050,075,125,150}/` (cn=1.0 reuses `generated/sdxl_canny/`) | ~30 min | `experiments/analyze_solutions.py` |
| **App C.8** LoRA fine-tune | `renders_textured/` + base DAv2-S model | ~1 h training + ~5 min eval | `experiments/finetune/{create_split,train_single_gpu,eval_nyu}.py` (training wrapper); pre-trained adapter at `experiments/finetune/dav2_lora_adapter/` |
| **App F.1** Cross-dataset | NYU-v2 + iBims-1 from official sources (NOT redistributed) | ~30 min | `scripts/evaluate_cross_dataset.py` |
| **App F.2** Complexity-MDE | already in main 219 GB | ~5 min | included in `evaluate_igf.py` |
| **App F.3** 3D reconstruction | `renders_textured/` (30 rooms L2_full) + 3D-FRONT meshes | ~1 h | `experiments/run_appendix_F_experiments.py --section F3_reconstruction` |
| **App F.4** MDE domain gap | + NYU-v2 from official source | ~30 min | `experiments/run_appendix_F_experiments.py --section F4_mde_domain` |
---
## Pre-Computed Result JSONs
> If you only need to **verify** the paper's numbers (no recompute), the JSONs
> below are sufficient.
| Paper item | JSON path on HF | Key |
|:---|:---|:---|
| Table 3 (main) | `evaluation/igf_summary.json` | per-model aggregated means |
| Tab error_decomp | `evaluation/error_decomposition.json` | per-model AbsRel decomposition |
| Tab mde_ablation | `evaluation/mde_ablation_summary.json` | SDXL/Flux × 3 MDE backbones |
| Tab neg_prompt | `evaluation/neg_prompt_ablation.json` | per-model w/ vs w/o |
| Tab n1_resolution | `evaluation/n1_resolution_ablation.json` | SD 1.5 native vs upsampled |
| Tab n3_lvp_2d | `evaluation/n3_lvp_improved.json` | per-model 2D L_vp |
| Tab seed_ablation | `evaluation/seed_ablation.json` | Flux.1 across 3 seeds |
| Wilcoxon p-values | `evaluation/wilcoxon.json` | per-pair Bonferroni-corrected |
| FID + LPIPS | `evaluation/fid_lpips.json` | per-model perceptual baselines |
App F downstream JSONs are released with the paper supplement
(<https://anonymous.4open.science/r/IGF-Bench-Code>) under `experiments/results/`.
---
## Schema Reference
### `selected_rooms.json`
```json
{
"rooms": [
{
"scene_id": "00110bde-f580-40be-b8bb-88715b338a2a",
"room_id": "Bedroom-43072",
"room_type": "Bedroom",
"space_type": "Bedroom",
"area_m2": 12.3
},
...
]
}
```
The on-disk directory name for each room is `{scene_id}_{room_id}`.
### `evaluation/igf_summary.json`
A list of dicts, one per model and one for `GT_baseline`:
```json
[
{
"model": "GT_baseline",
"n_samples": 3600,
"l_plane_residual_mean": 0.056,
"l_ortho_mean": 0.166,
"l_edge_f1_mean": 0.606,
"l_vp_deg_mean": 1.79,
"blender_abs_rel_mean": 0.0,
"depthpro_abs_rel_mean": 0.056,
...
},
{
"model": "sdxl",
"n_samples": 3600,
"delta_l_plane_mean": 0.052,
"delta_l_ortho_mean": 0.061,
"delta_l_edge_mean": -0.413,
...
},
...
]
```
### `evaluation/igf_results.json`
A dict keyed by model (`gt_baseline`, `sd15`, …, `kolors`), each holding a list
of 3,600 per-view records:
```json
{
"gt_baseline": [
{
"room": "00110bde-..._Bedroom-43072",
"level": "L0_empty",
"view": "view_0",
"l_plane_residual": 0.045,
"l_plane_inlier_ratio": 0.84,
"l_ortho": 0.130,
"n_normals": 5,
"l_edge_f1": 0.612,
"l_edge_precision": 0.59,
"l_edge_recall": 0.64,
"l_vp_deg": 0.41,
"depthpro_abs_rel": 0.052,
"depthpro_rmse": 0.118,
"depthpro_si_rmse": 0.041,
"depthpro_delta125": 0.984,
...
},
...
],
"sdxl": [...],
...
}
```
### `experiments/finetune/dav2_lora_adapter/adapter_config.json`
```json
{
"base_model_name_or_path": "depth-anything/Depth-Anything-V2-Metric-Indoor-Small-hf",
"peft_type": "LORA",
"r": 16,
"lora_alpha": 32,
"lora_dropout": 0.05,
"target_modules": ["query", "key", "value"],
"bias": "none"
}
```
Loadable via `peft.PeftModel.from_pretrained(base, adapter_dir)`.
---
## License Chain
```
IGF-Bench dataset
├── License: CC BY-NC-SA 4.0
│ • Inherits NC clause from upstream 3D-FRONT (Alibaba Tianchi NC license).
│ • Attribution required, share-alike, non-commercial use only.
├── Code (separate): Apache 2.0
│ • Located in the supplementary code repo, not on this dataset HF repo.
├── Per-asset upstream model licenses (binding for downstream redistribution):
│ ├── SD 1.5 → CreativeML Open RAIL-M
│ ├── SDXL 1.0 → CreativeML Open RAIL++-M
│ ├── SD 3.5 Large → Stability AI Community License (NC if revenue ≤ $1M/yr)
│ ├── Flux.1 Dev → FLUX.1 [dev] Non-Commercial License
│ ├── Hunyuan-DiT → Tencent Hunyuan Community License
│ ├── Kolors → Apache 2.0 + Kwai commercial-registration requirement
│ ├── DepthPro → Apple Sample Code License (apple-amlr; output redistribution
│ │ in a "gray area" — see DATASHEET §6 for full disclosure)
│ ├── Depth-Anything V2-Small (metric-indoor) → Apache 2.0
│ └── ZoeDepth → MIT
└── Eval-only datasets (NOT redistributed by us; obtain from the official source):
├── NYU-v2 → https://cs.nyu.edu/~silberman/datasets/nyu_depth_v2.html
└── iBims-1 → https://www.cvg.cit.tum.de/data/datasets/ibims1
```
For the full per-asset table see [`DATASHEET.md`](DATASHEET.md) §6
(Distribution).
The `dav2_lora_adapter/` weights inherit CC BY-NC-SA 4.0 from the 3D-FRONT
supervision data, even though the base DAv2-S model is Apache 2.0.
---
## Maintenance
The authors commit to maintaining IGF-Bench for **at least 5 years
post-publication**, including:
- Hosting on HuggingFace with versioned releases (`v1.0.0`, `v1.1.0`, …);
- Bug fixes via the GitHub Issues tracker (URL pending de-anonymisation);
- Adding new generation models as they become available;
- Periodic re-evaluation when major MDE backbones are released.
Older versions remain accessible on HuggingFace forever (see the
`refs/convert/parquet/<commit>` history).
---
## Citation
```bibtex
@inproceedings{igfbench2026,
title = {IGF-Bench: Evaluating Geometric Fidelity of Conditional Image
Generation Beyond Perceptual Metrics},
author = {Anonymous},
booktitle = {Advances in Neural Information Processing Systems
(Datasets and Benchmarks Track)},
year = {2026}
}
```
When citing the upstream 3D-FRONT scenes, please also cite Fu et al., *3D-FRONT: 3D Furnished Rooms with layOuts and semaNTics* (ICCV 2021).