Research map / 3D vision / 3D generation
DreamScene: 3D Gaussian-based Text-to-3D Scene Generation via Formation Pattern Sampling
Haoran Li, Haolin Shi, Wenli Zhang, Wenjun Wu, Yong Liao, Lin Wang, Lik-Hang Lee, Peng Yuan Zhou
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 |
|---|---|---|---|
| replace SDS and monotonic non-increasing sampling with multi-timestep sampling (MTS) | DreamScene (Iron Man, 30 min optimization) · single-object text-to-3D | qualitative geometry and texture | Improved |
| add reconstructive generation on top of MTS (full FPS) | DreamScene · single-object text-to-3D | qualitative texture | Improved |
| apply 3D Gaussian filtering in reconstruction task (ablation) | 3D Gaussian filtering · reconstruction task | image quality: 73.9% compression | Hurt |
| apply 3D Gaussian filtering during generation (ablation) | DreamScene · generation task | image quality: 66.1% compression | No clear effect |
| use three-stage camera sampling instead of random camera sampling | DreamScene · scene generation | optimization stability | Improved |
| separate environment and ground stages in camera sampling | DreamScene camera sampling · scene generation | ground-scene connection quality | Improved |
| DreamScene vs Set-the-Scene on generation time | DreamScene · scene generation (environment stage) | time (hours): 1.5 -> 1.0 | Improved |
| DreamScene vs ProlificDreamer on consistency | DreamScene · User study | consistency score: 3.19 -> 4.24 | Improved |
Later papers that compared with this method
4 later accepted papers report DreamScene as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- FilmSceneDesigner: Chaining Set Design for Procedural Film Scene Generation (AAAI 2026 · better on 8 of 8 shared results)
- RoomPlanner: Reachability-Aware View Sampling for Text-to-Room 3D Gaussian Splatting (ECCV 2026 · better on 5 of 5 shared results)
- Debiasing Diffusion Priors via 3D Attention for Consistent Gaussian Splatting (AAAI 2026 · better on 4 of 4 shared results)
- MagicCity: Geometry-Aware 3D City Generation from Satellite Imagery with Multi-View Consistency (ICCV 2025 · better on 6 of 6 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
- BloomScene: Lightweight Structured 3D Gaussian Splatting for Crossmodal Scene Generation (AAAI 2025 · similarity 0.89)
- ArtiScene: Language-Driven Artistic 3D Scene Generation Through Image Intermediary (CVPR 2025 · similarity 0.88)
- DreamControl: Control-Based Text-to-3D Generation with 3D Self-Prior (CVPR 2024 · similarity 0.88)
- GALA3D: Towards Text-to-3D Complex Scene Generation via Layout-guided Generative Gaussian Splatting (ICML 2024 · similarity 0.88)
- DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation (ICLR 2024 · similarity 0.88)
- 3D-SceneDreamer: Text-Driven 3D-Consistent Scene Generation (CVPR 2024 · similarity 0.88)
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.
