Can Vision Language Models Assess Graphic Design Aesthetics? A Benchmark, Evaluation, and Dataset Perspective

TL;DR

This paper introduces AesEval-Bench, a comprehensive benchmark and dataset for evaluating vision language models' ability to assess graphic design aesthetics, addressing previous limitations and providing a systematic framework.

Contribution

It presents the first systematic framework and benchmark for aesthetic assessment in graphic design using vision language models, including a new dataset and evaluation protocols.

Findings

VLMs show performance gaps in aesthetic judgment tasks.

Fine-tuning improves VLMs' assessment capabilities.

Benchmark reveals strengths and weaknesses of different VLMs.

Abstract

Assessing the aesthetic quality of graphic design is central to visual communication, yet remains underexplored in vision language models (VLMs). We investigate whether VLMs can evaluate design aesthetics in ways comparable to humans. Prior work faces three key limitations: benchmarks restricted to narrow principles and coarse evaluation protocols, a lack of systematic VLM comparisons, and limited training data for model improvement. In this work, we introduce AesEval-Bench, a comprehensive benchmark spanning four dimensions, twelve indicators, and three fully quantifiable tasks: aesthetic judgment, region selection, and precise localization. Then, we systematically evaluate proprietary, open-source, and reasoning-augmented VLMs, revealing clear performance gaps against the nuanced demands of aesthetic assessment. Moreover, we construct a training dataset to fine-tune VLMs for this domain, leveraging human-guided VLM labeling to produce task labels at scale and indicator-grounded reasoning to tie abstract indicators to concrete design regions.Together, our work establishes the first systematic framework for aesthetic quality assessment in graphic design. Our code and dataset will be released at: \href{https://github.com/arctanxarc/AesEval-Bench}{https://github.com/arctanxarc/AesEval-Bench}