PulseAugur
EN
LIVE 17:04:38

New frameworks advance 3D scene generation and editing

By PulseAugur Editorial · [4 sources] ·

Researchers have developed new frameworks for generating and editing 3D scenes from single images. SceneConductor decomposes the process into initialization, environment construction, and multi-agent refinement, improving geometric accuracy and realism. SceneCompleter focuses on dense 3D scene completion using a geometry-aware diffusion model for novel view synthesis, addressing issues of geometry distortion and appearance drift. MUSE offers agentic 3D scene authoring through memory-grounded incremental requirement satisfaction, enabling controllable construction and preservation-aware editing. AI

IMPACT These advancements in 3D scene generation and editing could accelerate content creation for virtual environments and AI simulations.

RANK_REASON The cluster contains multiple academic papers detailing new research frameworks for 3D scene generation and editing.

Read on arXiv cs.MA (Multiagent) →

AI-generated summary · Google Gemini · from 4 sources. How we write summaries →

New frameworks advance 3D scene generation and editing

COVERAGE [4]

  1. arXiv cs.MA (Multiagent) TIER_1 English(EN) · Xingang Pan ·

    SceneConductor: 3D Scene Generation from a Single Image with Multi-Agent Orchestration

    Generating complete 3D scenes from a single image requires inferring globally consistent geometry, object relationships, and environmental context from inherently ambiguous visual evidence. Despite recent progress in joint layout-and-mesh generation, existing methods often rely o…

  2. arXiv cs.CV TIER_1 English(EN) · Weiliang Chen, Jiayi Bi, Yuanhui Huang, Wenzhao Zheng, Yueqi Duan ·

    SceneCompleter: Dense 3D Scene Completion for Generative Novel View Synthesis

    arXiv:2506.10981v2 Announce Type: replace Abstract: Generative models have shown great promise for novel view synthesis (NVS) by leveraging strong image generation priors. However, existing approaches typically follow a 2D inpainting paradigm, first completing missing image regio…

  3. arXiv cs.CV TIER_1 English(EN) · Ruijie Xu, Xinnan Zhu, Jiayu Ying, Daoguo Dong, Yuzhou Ji, Xin Tan ·

    MUSE: Agentic 3D Scene Authoring via Memory-Grounded Incremental Requirement Satisfaction

    arXiv:2606.14168v1 Announce Type: new Abstract: Text-driven 3D scene generation is a promising technique for digital content creation, embodied AI simulation, and interactive design, yet practical workflows often require refining, extending, or correcting existing scenes while pr…

  4. arXiv cs.CV TIER_1 English(EN) · Xin Tan ·

    MUSE: Agentic 3D Scene Authoring via Memory-Grounded Incremental Requirement Satisfaction

    Text-driven 3D scene generation is a promising technique for digital content creation, embodied AI simulation, and interactive design, yet practical workflows often require refining, extending, or correcting existing scenes while preserving non-target content. Existing methods ca…

RELATED ENTITIES

RELATED TOPICS