CArtBench: Evaluating Vision-Language Models on Chinese Art Understanding, Interpretation, and Authenticity
This benchmark addresses the lack of fine-grained evaluation for VLMs on specialized cultural heritage tasks, revealing significant gaps in connoisseur-level reasoning.
CArtBench evaluates vision-language models on Chinese art understanding, interpretation, and authenticity. Results show that models struggle with evidence linking, style-to-period inference, and authenticity discrimination, often near chance.
We introduce CARTBENCH, a museum-grounded benchmark for evaluating vision-language models (VLMs) on Chinese artworks beyond short-form recognition and QA. CARTBENCH comprises four subtasks: CURATORQA for evidence-grounded recognition and reasoning, CATALOGCAPTION for structured four-section expert-style appreciation, REINTERPRET for defensible reinterpretation with expert ratings, and CONNOISSEURPAIRS for diagnostic authenticity discrimination under visually similar confounds. CARTBENCH is built by aligning image-bearing Palace Museum objects from Wikidata with authoritative catalog pages, spanning five art categories across multiple dynasties. Across nine representative VLMs, we find that high overall CURATORQA accuracy can mask sharp drops on hard evidence linking and style-to-period inference; long-form appreciation remains far from expert references; and authenticity-oriented diagnostic discrimination stays near chance, underscoring the difficulty of connoisseur-level reasoning for current models.