Composable scene modeling aims to recover a real indoor scene as complete, editable object assets arranged as observed, providing robot simulation and embodied AI with a simulation-ready replica of the real environment where objects can be manipulated individually. Existing pipelines decompose the task into three steps: parse the observations into instances, generate an asset for each, and place each asset back. However, every step presumes an input that a cluttered capture rarely provides: accurate instance geometry, unoccluded views, and assets that accurately match the observations.
To address this, researchers propose Lucida, which keeps the same order but redistributes the requirements so that each step consumes only what a real capture reliably provides, and precision is reached at the end of the pipeline rather than demanded at its start. Lucida parses the video into a scene graph whose nodes carry per-instance multi-view evidence, generates a complete asset for each instance from its evidence, and places assets with GizmoAct, a VLM policy that casts placement as multi-turn GUI interaction, manipulating the object's gizmo in a closed loop and deciding itself when alignment is reached.
Across scene-level 3D object detection, object pose estimation, and scene reconstruction, Lucida improves mAP over Boxer by 69% on R2S-Scene, raises [email protected] from 57.8% to 83.4% on CA-1M, and increases scene F-Score from 0.794 for SAM3D to 0.924.