arXiv：cs.AI· Ivan Yuri De Leon, Hiroki Ouchi, Sakriani Sakti·· 4 小时前
Tanawpamana：菲律宾文化遗产地视觉语言模型基准测试
Tanawpamana: Benchmarking Vision-Language Models on Philippine Cultural Heritage Sites
AI 导读
研究团队推出 Tanawpamana 数据集与基准，包含 4,046 项任务，用于评估视觉语言模型对菲律宾文化遗产地的识别与知识。测试七个视觉语言模型发现，它们在相似遗址的细粒度视觉区分上比粗略识别更吃力，且从图像输入推断地理知识的表现弱于文本输入。
正文
Abstract:Vision-language models are increasingly used to identify and describe places, yet their recognition and knowledge of specific cultures remain poorly understood, particularly in underrepresented regions such as the Philippines. This study introduces the Tanawpamana Dataset, a repository of Philippine heritage site images with metadata based on the Talapamana registry of the National Commission for Culture and the Arts of the Philippines. We also introduce the Tanawpamana Benchmark, comprising 4,046 tasks designed to evaluate models' recognition and knowledge of heritage sites. Evaluating seven vision-language models, we find that they struggle more with fine-grained visual discrimination among similar heritage sites than with coarse recognition, and that their geographic knowledge is weaker when inferred from image inputs than from text.
| Comments: | Accepted at O-COCOSDA 2026 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.17596 [cs.CV] |
| (or arXiv:2610.17596v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.17596 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ivan Yuri De Leon [view email]
[v1]
Tue, 13 Oct 2026 02:15:52 UTC (1,724 KB)
来源：arXiv：cs.AI · arxiv.org