Metraipolis › archive › dc_0cb63ca33ba2
TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows — arXiv:2610.02959
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.02959
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T09:07:31.374Z
- SHA-256 of the archived text
e12e4aee089f634cbe1ff501c405f492f586e27a844f333ccf295cba94b13806- Stable URL of this page
https://metraipolis.com/documento/dc_0cb63ca33ba2
How to cite
TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows — arXiv:2610.02959 — arXiv (en), https://arxiv.org/abs/2610.02959 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0cb63ca33ba2 (Metraipolis archive dc_0cb63ca33ba2, sha256 e12e4aee089f634cbe1ff501c405f492f586e27a844f333ccf295cba94b13806)
BibTeX
@misc{dc0cb63ca33ba2,
title = {TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows — arXiv:2610.02959},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.02959},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0cb63ca33ba2; sha256 e12e4aee089f634cbe1ff501c405f492f586e27a844f333ccf295cba94b13806},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.02959. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recent text-to-image models have made substantial progress in photorealism, aesthetics, and text-image alignment. Yet visually appealing images can still violate real-world plausibility, exhibiting malformed object structures, impossible anatomy, physically implausible interactions, or inconsistent spatial relationships. Such failures are not well captured by existing fidelity, aesthetics, preference, or alignment metrics. To address this gap, we introduce TerraVis, a framework for evaluating world-grounded visual consistency in generated images. TerraVis defines a structured taxonomy of world-consistency violations spanning object-, interaction-, and scene-level failures, and employs a multi-stage evaluation framework to identify and quantify them. Given an image, TerraVis first uses an MLLM to assess its eligibility for evaluation, then detects violations across 18 taxonomy-defined types and classifies them as minor or major to derive an overall world-consistency score. Across diverse open-source and proprietary text-to-image models on two widely used benchmarks, TerraVis achieves the strongest correlation with human judgments of world consistency among existing metrics. Our benchmark results further show that models that achieve strong performance on conventional metrics can still exhibit substantial world-consistency failures. These findings highlight world consistency as a complementary evaluation dimension and demonstrate that TerraVis enables systematic quantification, diagnosis, and comparison of such failures. Our code is publicly available at this https URL . Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-05 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2610.02959 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-06T08:21:44.190Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-05?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).