SceneTeract: Agentic Functional Affordances and VLM Grounding in 3D Scenes

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Maillard, Léopold, Engelmann, Francis, Durand, Tom, Pan, Boxiao, You, Yang, Litany, Or, Guibas, Leonidas, Ovsjanikov, Maks
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911558030852096
author Maillard, Léopold
Engelmann, Francis
Durand, Tom
Pan, Boxiao
You, Yang
Litany, Or
Guibas, Leonidas
Ovsjanikov, Maks
author_facet Maillard, Léopold
Engelmann, Francis
Durand, Tom
Pan, Boxiao
You, Yang
Litany, Or
Guibas, Leonidas
Ovsjanikov, Maks
contents Embodied AI depends on interactive 3D environments that support meaningful activities for diverse users, yet assessing their functional affordances remains a core challenge. We introduce SceneTeract, a framework that verifies 3D scene functionality under agent-specific constraints. Our core contribution is a grounded verification engine that couples high-level semantic reasoning with low-level geometric checks. SceneTeract decomposes complex activities into sequences of atomic actions and validates each step against accessibility requirements (e.g., reachability, clearance, and navigability) conditioned on an embodied agent profile, using explicit physical and geometric simulations. We deploy SceneTeract to perform an in-depth evaluation of (i) synthetic indoor environments, uncovering frequent functional failures that prevent basic interactions, and (ii) the ability of frontier Vision-Language Models (VLMs) to reason about and predict functional affordances, revealing systematic mismatches between semantic confidence and physical feasibility even for the strongest current models. Finally, we leverage SceneTeract as a reward engine for VLM post-training, enabling scalable distillation of geometric constraints into reasoning models. We release the SceneTeract verification suite and data to bridge perception and physical reality in embodied 3D scene understanding.
format Preprint
id arxiv_https___arxiv_org_abs_2603_29798
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SceneTeract: Agentic Functional Affordances and VLM Grounding in 3D Scenes
Maillard, Léopold
Engelmann, Francis
Durand, Tom
Pan, Boxiao
You, Yang
Litany, Or
Guibas, Leonidas
Ovsjanikov, Maks
Computer Vision and Pattern Recognition
Embodied AI depends on interactive 3D environments that support meaningful activities for diverse users, yet assessing their functional affordances remains a core challenge. We introduce SceneTeract, a framework that verifies 3D scene functionality under agent-specific constraints. Our core contribution is a grounded verification engine that couples high-level semantic reasoning with low-level geometric checks. SceneTeract decomposes complex activities into sequences of atomic actions and validates each step against accessibility requirements (e.g., reachability, clearance, and navigability) conditioned on an embodied agent profile, using explicit physical and geometric simulations. We deploy SceneTeract to perform an in-depth evaluation of (i) synthetic indoor environments, uncovering frequent functional failures that prevent basic interactions, and (ii) the ability of frontier Vision-Language Models (VLMs) to reason about and predict functional affordances, revealing systematic mismatches between semantic confidence and physical feasibility even for the strongest current models. Finally, we leverage SceneTeract as a reward engine for VLM post-training, enabling scalable distillation of geometric constraints into reasoning models. We release the SceneTeract verification suite and data to bridge perception and physical reality in embodied 3D scene understanding.
title SceneTeract: Agentic Functional Affordances and VLM Grounding in 3D Scenes
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.29798