Female Identity and Class Imagination in Qing Chinese Export Paintings: An AI-Assisted Study of Cross-Cultural Interpretation

This study investigated female identity and class imagination in Qing Chinese export paintings through an AI-assisted cross-cultural interpretation framework. Methods: A corpus of 433 export paintings dated from 1757 to 1911 was constructed from museum collections, digital archives, and published catalogues. CLIP-based visual embedding, segmentation-assisted inspection, UMAP clustering, and manual iconographic coding were combined to classify female identity types and class-marking visual symbols. Results: Seven recurrent female identity categories were identified: elite/beauty women (n = 112), laboring women (n = 72), domestic women (n = 64), servant women (n = 58), ritual/wedding women (n = 45), market women (n = 43), and performing women (n = 39). AI–human agreement reached 82.2%, with the highest agreement for elite/beauty women (88.4%) and the lowest for ritual/wedding women (73.3%). Conclusion: Female identity in Qing export paintings was visually stabilized through class-coded markers and mediated by cross-cultural market consumption.