Research map / 3D vision / 3D generation
UrbanGIRAFFE: Representing Urban Scenes as Compositional Generative Neural Feature Fields
Yuanbo Yang, Yifei Yang, Hanlei Guo, Rong Xiong, Yue Wang, Yiyi Liao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add stuff reconstruction loss (MSE plus perceptual) on non-object regions (ablation) | UrbanGIRAFFE · KITTI-360 | FID_I: 89.3 -> 39.0 | Improved |
| add object-level discriminator on object patches (ablation) | UrbanGIRAFFE · KITTI-360 | FID_P (object patches): 119.0 -> 67.1 | Improved |
| add object-level discriminator on object patches (ablation) | UrbanGIRAFFE · KITTI-360 | FID_I: 53.1 -> 39.0 | Improved |
| add separate object generator with canonical-space objects instead of stuff generator for whole scene (ablation) | UrbanGIRAFFE · KITTI-360 | FID_I: 44.7 -> 39.0 | Improved |
| add separate object generator (ablation) | UrbanGIRAFFE · KITTI-360 | KID_I: 0.036 -> 0.036 | No clear effect |
| use semantic-voxel conditioned stuff generator and panoptic prior vs GIRAFFE background | UrbanGIRAFFE · KITTI-360 | FID_I: 112.1 -> 39.6 | Improved |
| 3D-aware panoptic-conditioned model vs 2D StyleGAN2 baseline | UrbanGIRAFFE · KITTI-360 | FID_I: 31.9 (2D GAN) vs 39.6 | Hurt |
| UrbanGIRAFFE vs GIRAFFE on CLEVR-W | UrbanGIRAFFE · CLEVR-W | FID_I: 103.9 -> 25.7 | Improved |
Later papers that compared with this method
1 later accepted paper reports UrbanGIRAFFE as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Generative Gaussian Splatting for Unbounded 3D City Generation (CVPR 2025 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GIRAFFE — GIRAFFE: Representing Scenes As Compositional Generative Neural Feature Fields (CVPR 2021) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- UrbanGen: Urban Generation with Compositional and Controllable Neural Fields (TPAMI 2025 · similarity 0.91)
- PrITTI: Primitive-based Generation of Controllable and Editable 3D Semantic Urban Scenes (CVPR 2026 · similarity 0.90)
- CityDreamer: Compositional Generative Model of Unbounded 3D Cities (CVPR 2024 · similarity 0.90)
- CymbaDiff: Structured Spatial Diffusion for Sketch-based 3D Semantic Urban Scene Generation (NeurIPS 2025 · similarity 0.87)
- Dynamic 3D Gaussian Fields for Urban Areas (NeurIPS 2024 · similarity 0.87)
- Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata (CVPR 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
