PalmLeaf-VQA：多文字历史贝叶斯手稿视觉问答基准
原标题：PalmLeaf-VQA: A Multi-Script Visual Question Answering Benchmark for Historical Palm-Leaf Manuscript Understanding Across Diverse Regions
AI 摘要
研究者提出 PalmLeaf-VQA，一个面向南亚与东南亚历史贝叶斯手稿理解的多文字视觉问答基准，包含 923 张精选图像和 7,384 个新标注问答对，覆盖八个收藏组和八类手稿感知任务。该基准采用图像不重叠的训练/验证/测试划分，并对近期闭源与开源多模态大模型进行零样本评测，最强模型在测试集上仅达 58.00% 精确匹配准确率，显示当前 MLLM 在稀有文字、退化版面和保存相关文档理解上仍有明显不足。
正文节选
PalmLeaf-VQA: A Multi-Script Visual Question Answering Benchmark for Historical Palm-Leaf Manuscript Understanding Across Diverse Regions Abstract Historical manuscripts remain largely absent from modern vision-language benchmarks, leaving open how well multimodal large language models (MLLMs) handle culturally diverse, degraded, and non-Latin document images. We introduce PalmLeaf-VQA, a multi-script visual question answering benchmark for historical palm-leaf manuscript understanding across So