DesignSense: A Human Preference Dataset and Reward Modeling Framework for Graphic Layout Generation

TL;DR

This paper introduces DesignSense, a large-scale human preference dataset and a reward modeling framework for graphic layout generation, addressing the gap in aligning models with nuanced human aesthetic judgments.

Contribution

The paper presents a novel dataset, DesignSense-10k, and a specialized reward model that improves layout evaluation and generation quality over existing models.

Findings

DesignSense-10k contains 10,235 human-annotated preference pairs.

The reward model outperforms existing models with a 54.6% improvement in Macro F1.

Using the model in RL training improves layout generator win rate by 3%.

Abstract

Graphic layouts serve as an important and engaging medium for visual communication across different channels. While recent layout generation models have demonstrated impressive capabilities, they frequently fail to align with nuanced human aesthetic judgment. Existing preference datasets and reward models trained on text-to-image generation do not generalize to layout evaluation, where the spatial arrangement of identical elements determines quality. To address this critical gap, we introduce DesignSense-10k, a large-scale dataset of 10,235 human-annotated preference pairs for graphic layout evaluation. We propose a five-stage curation pipeline that generates visually coherent layout transformations across diverse aspect ratios, using semantic grouping, layout prediction, filtering, clustering, and VLM-based refinement to produce high-quality comparison pairs. Human preferences are annotated using a 4-class scheme (left, right, both good, both bad) to capture subjective ambiguity. Leveraging this dataset, we train DesignSense, a vision-language model-based classifier that substantially outperforms existing open-source and proprietary models across comprehensive evaluation metrics (54.6% improvement in Macro F1 over the strongest proprietary baseline). Our analysis shows that frontier VLMs remain unreliable overall and fail catastrophically on the full four-class task, underscoring the need for specialized, preference-aware models. Beyond the dataset, our reward model DesignSense yields tangible downstream gains in layout generation. Using our judge during RL based training improves generator win rate by about 3%, while inference-time scaling, which involves generating multiple candidates and selecting the best one, provides a 3.6% improvement. These results highlight the practical impact of specialized, layout-aware preference modeling on real-world layout generation quality.