DreamScene : 3D Gaussian-based End-to-end Text-to-3D Scene Generation

Generating 3D scenes from natural language holds great promise for applications in gaming, film, and design. However, existing methods struggle with automation, 3D consistency, and fine-grained control. We present DreamScene, an end-to-end framework for high-quality and editable 3D scene generation...

Ausführliche Beschreibung

Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on pattern analysis and machine intelligence. - 1979. - PP(2025) vom: 15. Okt.
1. Verfasser: Li, Haoran (VerfasserIn)
Weitere Verfasser: Tian, Yuli, Lan, Kun, Liao, Yong, Wang, Lin, Hui, Pan, Zhou, Peng Yuan
Format: Online-Aufsatz
Sprache:English
Veröffentlicht: 2025
Zugriff auf das übergeordnete Werk:IEEE transactions on pattern analysis and machine intelligence
Schlagworte:Journal Article
LEADER 01000naa a22002652c 4500
001 NLM394037987
003 DE-627
005 20251017232232.0
007 cr uuu---uuuuu
008 251017s2025 xx |||||o 00| ||eng c
024 7 |a 10.1109/TPAMI.2025.3620403  |2 doi 
028 5 2 |a pubmed25n1601.xml 
035 |a (DE-627)NLM394037987 
035 |a (NLM)41091609 
040 |a DE-627  |b ger  |c DE-627  |e rakwb 
041 |a eng 
100 1 |a Li, Haoran  |e verfasserin  |4 aut 
245 1 0 |a DreamScene  |b 3D Gaussian-based End-to-end Text-to-3D Scene Generation 
264 1 |c 2025 
336 |a Text  |b txt  |2 rdacontent 
337 |a ƒaComputermedien  |b c  |2 rdamedia 
338 |a ƒa Online-Ressource  |b cr  |2 rdacarrier 
500 |a Date Revised 15.10.2025 
500 |a published: Print-Electronic 
500 |a Citation Status Publisher 
520 |a Generating 3D scenes from natural language holds great promise for applications in gaming, film, and design. However, existing methods struggle with automation, 3D consistency, and fine-grained control. We present DreamScene, an end-to-end framework for high-quality and editable 3D scene generation from text or dialogue. DreamScene begins with a scene planning module, where a GPT-4 agent infers object semantics and spatial constraints to construct a hybrid graph. A graph-based placement algorithm then produces a structured, collision-free layout. Based on this layout, Formation Pattern Sampling (FPS) generates object geometry using multi-timestep sampling and reconstructive optimization, enabling fast and realistic synthesis. To ensure global consistent, DreamScene employs a progressive camera sampling strategy tailored to both indoor and outdoor settings. Finally, the system supports fine-grained scene editing, including object movement, appearance changes, and 4D dynamic motion. Experiments demonstrate that DreamScene surpasses prior methods in quality, consistency, and flexibility, offering a practical solution for open-domain 3D content creation. Code and demos are available at https://dreamscene-project.github.io 
650 4 |a Journal Article 
700 1 |a Tian, Yuli  |e verfasserin  |4 aut 
700 1 |a Lan, Kun  |e verfasserin  |4 aut 
700 1 |a Liao, Yong  |e verfasserin  |4 aut 
700 1 |a Wang, Lin  |e verfasserin  |4 aut 
700 1 |a Hui, Pan  |e verfasserin  |4 aut 
700 1 |a Zhou, Peng Yuan  |e verfasserin  |4 aut 
773 0 8 |i Enthalten in  |t IEEE transactions on pattern analysis and machine intelligence  |d 1979  |g PP(2025) vom: 15. Okt.  |w (DE-627)NLM098212257  |x 1939-3539  |7 nnas 
773 1 8 |g volume:PP  |g year:2025  |g day:15  |g month:10 
856 4 0 |u http://dx.doi.org/10.1109/TPAMI.2025.3620403  |3 Volltext 
912 |a GBV_USEFLAG_A 
912 |a SYSFLAG_A 
912 |a GBV_NLM 
912 |a GBV_ILN_350 
951 |a AR 
952 |d PP  |j 2025  |b 15  |c 10