TensorX
返回文献探索

Paper · arXiv 2608.30821

Lucida: Parse, Generate, and Place for Composable Real-to-Sim Scene Modeling

Minghan Qin, Yuang Wang, Xiuyu Yang, Yushi Long, Yujian Zhang, Ruihuan Wang, Kai Ye, Yangang Zhang, Hang Li

119 upvotesAugust 31, 2026arXiv 预印本
AI 摘要

Lucida improves composable indoor scene reconstruction by distributing pipeline requirements across parsing, asset generation, and VLM-guided placement to achieve high-fidelity editable replicas from cluttered captures.

scene graphmulti-view evidenceVLM policyGizmoActmulti-turn GUI interaction3D object detectionobject pose estimationscene reconstructionADD-SBF-Score

Abstract

Composable scene modeling aims to recover a real indoor scene as complete, editable object assets arranged as observed, giving robot simulation and embodied AI a simulation-ready replica of the real environment whose 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---but every step presumes an input that a cluttered capture rarely provides: accurate instance geometry, unoccluded views, and assets that accurately match the observations. We propose Lucida, which keeps this order but redistributes the requirements, so 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 ADD-SB@0.05 from 57.8% to 83.4% on CA-1M, and increases scene F-Score from 0.794 for SAM3D to 0.924.

北京市昌平区探索星信息技术及软件开发工作室

京ICP备2026059466号