XPASS-Vis: A Dataset for Cross-Domain Personalized Image Aesthetic Assessment

2026-06-14

AI Summary

This study addresses the limitation of existing personalized image aesthetic assessment research, which is confined to a single visual domain and lacks large-scale annotated data necessary for cross-domain modeling. To bridge this gap, we introduce XPASS-Vis, the first cross-domain personalized aesthetic dataset encompassing art, fashion, and landscape images, with each user providing annotations for over 200 images per domain. Leveraging this dataset, we propose an unsupervised domain adaptation (UDA) approach and demonstrate, for the first time, the cross-domain transferability of personalized aesthetic preferences. Experimental results show that under a fully unsupervised setting, the best-performing UDA model achieves approximately 60% of the performance of a supervised upper bound (Spearman’s ρ = 0.28), establishing the feasibility of cross-domain personalized aesthetic assessment.

Abstract

Personalized image aesthetic assessment (PIAA) seeks to model, at the individual level, the subjective nature of aesthetic judgments toward artworks and photographs. Aesthetic preference is known to be both deeply personal and partially consistent across visual domains. Yet existing PIAA datasets and methods are largely confined to a single domain, or provide too few samples per annotator within each domain to enable personalization across domains. Consequently, the cross-domain generalization of personalized aesthetic preferences remains largely unexplored. To address this gap, we introduce XPASS-Vis, the first dataset explicitly designed for cross-domain PIAA. XPASS-Vis comprises 6,526 stimuli from three visual domains -- art, fashion, and landscape -- rated by 129 annotators, yielding 87,836 user-stimulus interactions, each annotated with an overall aesthetic score and nine aesthetic-emotion ratings. Notably, each annotator rated more than 200 stimuli per domain, providing sufficient per-domain coverage to support personalization both within and across domains. Moreover, we establish baseline models for cross-domain PIAA under unsupervised domain adaptation (UDA), where a model trained on a labeled source domain is transferred to an unlabeled target domain. A systematic evaluation of representative UDA approaches shows that the best-performing method recovers approximately 60\% (Spearman's $ρ$ = .28) of the supervised upper bound under a fully unsupervised setting. This provides encouraging evidence that personalized aesthetic preferences are, to a meaningful extent, transferable across visual domains. At the same time, a substantial gap remains, highlighting the need for PIAA-specific adaptation strategies. XPASS-Vis and the accompanying baselines provide a foundation for future research on cross-domain PIAA. All datasets and code will be made publicly available upon acceptance.

Problem

• Personalized Image Aesthetic Assessment
• Cross-Domain Generalization
• Aesthetic Preference
• Visual Domains
• User-Subjective Judgment

Innovation

• cross-domain personalization
• image aesthetic assessment
• unsupervised domain adaptation
• aesthetic-emotion ratings
• XPASS-Vis dataset