SceneCritic: A Symbolic Evaluator for 3D Indoor Scene Synthesis

2026-04-14

AI Summary

Existing 3D indoor scene evaluation methods based on large language models (LLMs) or vision-language models (VLMs) are highly sensitive to viewpoint variations, prompt phrasing, and model hallucinations, leading to unstable assessments of spatial plausibility. To address this, this work proposes SceneCritic—a symbolic rule-based evaluation framework that constructs SceneOnto, a multi-source fused structured spatial ontology, to jointly verify semantic, orientational, and geometric relationships among objects, enabling fine-grained violation detection and iterative refinement. Experiments demonstrate that SceneCritic achieves significantly higher agreement with human judgments than VLMs; furthermore, text-only LLMs can outperform VLMs in semantic layout evaluation, while VLMs exhibit the strongest correction capability when provided with visual feedback, collectively validating the complementary effectiveness of different critic modalities in optimizing spatial configurations.

Technology Category

• Computer Vision: Scene Analysis & Understanding
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Large Language Models (LLMs) and Vision-Language Models (VLMs) increasingly generate indoor scenes through intermediate structures such as layouts and scene graphs, yet evaluation still relies on LLM or VLM judges that score rendered views, making judgments sensitive to viewpoint, prompt phrasing, and hallucination. When the evaluator is unstable, it becomes difficult to determine whether a model has produced a spatially plausible scene or whether the output score reflects the choice of viewpoint, rendering, or prompt. We introduce SceneCritic, a symbolic evaluator for floor-plan-level layouts. SceneCritic's constraints are grounded in SceneOnto, a structured spatial ontology we construct by aggregating indoor scene priors from 3D-FRONT, ScanNet, and Visual Genome. SceneOnto traverses this ontology to jointly verify semantic, orientation, and geometric coherence across object relationships, providing object-level and relationship-level assessments that identify specific violations and successful placements. Furthermore, we pair SceneCritic with an iterative refinement test bed that probes how models build and revise spatial structure under different critic modalities: a rule-based critic using collision constraints as feedback, an LLM critic operating on the layout as text, and a VLM critic operating on rendered observations. Through extensive experiments, we show that (a) SceneCritic aligns substantially better with human judgments than VLM-based evaluators, (b) text-only LLMs can outperform VLMs on semantic layout quality, and (c) image-based VLM refinement is the most effective critic modality for semantic and orientation correction.

Problem

• 3D indoor scene synthesis
• scene evaluation
• spatial plausibility
• viewpoint sensitivity
• hallucination

Innovation

• symbolic evaluation
• spatial ontology
• 3D scene synthesis
• layout refinement
• scene graph validation