跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-02AI 評分44

TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows

TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows

AI 導讀

TerraVis 提出一套評估文生圖生成影像世界一致性的框架,透過多階段流程辨識物件、互動及場景層面的違規情況。框架涵蓋 18 種違規類型,並將嚴重程度分為輕微或重大;在兩個常用基準上,TerraVis 與人類世界一致性評分的相關性高於現有指標。研究亦顯示,模型即使在傳統指標上表現出色,仍可能出現明顯的世界一致性問題。

正文

Published on Oct 2

Authors:

,

,

,

,

Abstract

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 https://github.com/ShyFoo/TerraVis.

View arXiv page View PDF Project page GitHub 1 Add to collection

Get this paper in your agent:

hf papers read 2610.02959

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.02959 in a model README.md to link it from this page.

Datasets citing this paper 1

ShyFoo/TerraVis-Annotations

Viewer •

Updated about 1 hour ago

•

4.5k

•

19

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.02959 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co