一作ACL 2026Findings· Findings of ACL
Dunhuang-Bench: How Well Do MLLM s Understand Cultural Heritage?
Junyi Yuan · Jian Zhang · Tianxiu Yu · Yanlin Zhou · Xiaobo Jin · Qiufeng Wang · Fangyu Wu
摘要
Dunhuang art, a cornerstone of global heritage, demands fine-grained visual perception anchored by specialized cultural knowledge. Given the strong performance of multimodal large language models (MLLMs) on generic multimodal benchmarks, to what extent can they understand artifacts from Dunhuang...