DesignAgent3D: Interactive 3D Scene Editing via Designer-like Multimodal Reasoning

DesignAgent3D, an interactive multimodal agentic framework introduced in arXiv:2608.21438v1, reformulates 3D scene editing as a Plan-Perceive-Act paradigm to resolve ambiguous user intents and achieve precise spatial grounding. It outperforms state-of-the-art baselines across NeRF and 3D Gaussian Splatting backbones, delivering superior semantic intent alignment, impeccable spatial localization accuracy, and high-fidelity multi-view consistency.

arXiv:2608.21438v1 Announce Type: new Abstract: Text guided 3D scene editing provides an intuitive interface for modifying reconstructed environments, but remains difficult because natural language design requests are often semantically underspecified and must be grounded in cluttered 3D scenes. Existing methods typically formulate the task as one-shot conditional generation from a single prompt, failing to resolve ambiguous user intents or achieve precise spatial grounding. Consequently, they suffer from severe object localization drift, tracking failure under occlusions, and the notorious multi-view "sticker effect." To overcome these limitations, we present DesignAgent3D, an interactive multimodal agentic framework that reformulates 3D scene editing as a designer-like Plan-Perceive-Act paradigm. The agent first plans by interacting with the user to clarify underspecified design goals, then perceives by grounding the intended edit to specific objects or regions in the 3D scene, and finally acts by applying controlled visual modifications while preserving scene consistency. The edits are further integrated into the underlying 3D representation, supporting persistent and multi-view consistent novel-view rendering. Extensive experiments across both NeRF and 3D Gaussian Splatting backbones demonstrate that DesignAgent3D significantly outperforms state-of-the-art baselines, delivering superior semantic intent alignment, impeccable spatial localization accuracy, and high-fidelity multi-view consistency.