Cross-Platform Aesthetic Analysis: A Computer Vision Study of REDnote and Instagram

Visual aesthetic evaluation is a crucial social and cultural activity, influencing human visual communication and cultural exchange. While large-scale AI models offer new opportunities for cross-cultural aesthetic analysis, existing studies focus on visual message production, effects, or single cultural contexts, neglecting cross-cultural comparisons and emerging AI-generated content like digital humans. This study implements a cross-cultural approach, using representative large language models to evaluate Instagram and REDnote beauty influencers' photos and AI-generated digital human images based on universal portrait aesthetic criteria. Results reveal significant variations in aesthetic scores among the AI models. These findings underscore the potential of leveraging large-scale AI models for cross-cultural aesthetic studies. The models demonstrated limited capabilities in identifying digital humans. This research provides new insights for understanding cultural differences in machine-based aesthetic judgments.