Let Language Constrain Geometry: Vision-Language Models as Semantic and Spatial Critics for 3D Generation

TL;DR

VLM3D leverages large vision-language models as semantic and spatial critics to improve the accuracy and coherence of text-to-3D generation, addressing semantic detail and spatial consistency issues.

Contribution

It introduces a dual-query critic signal from VLMs for semantic and spatial evaluation, applicable to both optimization-based and feed-forward 3D generation methods.

Findings

Outperforms existing methods on standard benchmarks.

Effectively corrects spatial errors during 3D generation.

Enhances semantic fidelity in generated 3D models.

Abstract

Text-to-3D generation has advanced rapidly, yet state-of-the-art models, encompassing both optimization-based and feed-forward architectures, still face two fundamental limitations. First, they struggle with coarse semantic alignment, often failing to capture fine-grained prompt details. Second, they lack robust 3D spatial understanding, leading to geometric inconsistencies and catastrophic failures in part assembly and spatial relationships. To address these challenges, we propose VLM3D, a general framework that repurposes large vision-language models (VLMs) as powerful, differentiable semantic and spatial critics. Our core contribution is a dual-query critic signal derived from the VLM's Yes or No log-odds, which assesses both semantic fidelity and geometric coherence. We demonstrate the generality of this guidance signal across two distinct paradigms: (1) As a reward objective for optimization-based pipelines, VLM3D significantly outperforms existing methods on standard benchmarks. (2) As a test-time guidance module for feed-forward pipelines, it actively steers the iterative sampling process of SOTA native 3D models to correct severe spatial errors. VLM3D establishes a principled and generalizable path to inject the VLM's rich, language-grounded understanding of both semantics and space into diverse 3D generative pipelines.