Human-in-the-Loop Local Corrections of 3D Scene Layouts via Infilling

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xie, Christopher, Avetisyan, Armen, Howard-Jenkins, Henry, Siddiqui, Yawar, Straub, Julian, Newcombe, Richard, Balntas, Vasileios, Engel, Jakob
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912511304925184
author Xie, Christopher
Avetisyan, Armen
Howard-Jenkins, Henry
Siddiqui, Yawar
Straub, Julian
Newcombe, Richard
Balntas, Vasileios
Engel, Jakob
author_facet Xie, Christopher
Avetisyan, Armen
Howard-Jenkins, Henry
Siddiqui, Yawar
Straub, Julian
Newcombe, Richard
Balntas, Vasileios
Engel, Jakob
contents We present a novel human-in-the-loop approach to estimate 3D scene layout that uses human feedback from an egocentric standpoint. We study this approach through introduction of a novel local correction task, where users identify local errors and prompt a model to automatically correct them. Building on SceneScript, a state-of-the-art framework for 3D scene layout estimation that leverages structured language, we propose a solution that structures this problem as "infilling", a task studied in natural language processing. We train a multi-task version of SceneScript that maintains performance on global predictions while significantly improving its local correction ability. We integrate this into a human-in-the-loop system, enabling a user to iteratively refine scene layout estimates via a low-friction "one-click fix'' workflow. Our system enables the final refined layout to diverge from the training distribution, allowing for more accurate modelling of complex layouts.
format Preprint
id arxiv_https___arxiv_org_abs_2503_11806
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Human-in-the-Loop Local Corrections of 3D Scene Layouts via Infilling
Xie, Christopher
Avetisyan, Armen
Howard-Jenkins, Henry
Siddiqui, Yawar
Straub, Julian
Newcombe, Richard
Balntas, Vasileios
Engel, Jakob
Computer Vision and Pattern Recognition
We present a novel human-in-the-loop approach to estimate 3D scene layout that uses human feedback from an egocentric standpoint. We study this approach through introduction of a novel local correction task, where users identify local errors and prompt a model to automatically correct them. Building on SceneScript, a state-of-the-art framework for 3D scene layout estimation that leverages structured language, we propose a solution that structures this problem as "infilling", a task studied in natural language processing. We train a multi-task version of SceneScript that maintains performance on global predictions while significantly improving its local correction ability. We integrate this into a human-in-the-loop system, enabling a user to iteratively refine scene layout estimates via a low-friction "one-click fix'' workflow. Our system enables the final refined layout to diverge from the training distribution, allowing for more accurate modelling of complex layouts.
title Human-in-the-Loop Local Corrections of 3D Scene Layouts via Infilling
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.11806