pipette
ENEnglish

SceneRig: An Agentic System for Simulation-Ready 3D Scene Reconstruction from Single Images

Rundong Luo, Yunong Liu, Di Cao, Matthew Tancik, Shyamal Buch, Colton Stearns, Wenqi Xian

PreprintUso en el mundo real

En palabras de los autores

Simulation is a powerful tool for robotics, but building diverse, simulation-ready scenes at scale remains difficult. Reconstructing scenes from single images is a promising alternative, but visual similarity alone is insufficient: objects must also be recovered, accurately placed, and physically grounded. Existing methods that rely on post-hoc physics settling often misplace objects, producing scenes that look correct but behave poorly under manipulation. We present SceneRig, an agentic system that turns a single image into an editable, simulation-ready scene. SceneRig combines geometric, visual, and physical tools to reconstruct a faithful 3D scene. An agent then performs edit-and-validate loops inside a physics simulator to achieve physical stability. Across diverse inputs, SceneRig improves instance segmentation, object-centric appearance, and 3D placement while keeping the reconstructed scenes physically stable. Furthermore, SceneRig is effective for downstream robotics: its scenes predict real-robot policy outcomes more accurately than the baseline method and more reliably support replay of recorded manipulations without trajectory adaptation.

Resultado principalEl resumen no menciona limitaciones.

Apareció: miércoles, 21 de octubre. arXiv. Preprint, todavía sin revisión por pares.

Comentario de los autores: Accepted at CoRL 2026 Agentic Robotics Workshop. Project page: https://red-fairy.github.io/SceneRig/