Benchmarking Vision-Language Models on Chinese Ancient Documents: From OCR to Knowledge Reasoning
This addresses the problem of digitizing and understanding complex Chinese ancient documents for researchers and cultural preservationists, but it is incremental as it extends existing benchmark approaches to a new domain.
The authors tackled the lack of evaluation for Vision-Language Models on Chinese ancient documents by creating AncientDoc, the first benchmark covering tasks from OCR to knowledge reasoning, which includes over 3,000 pages and 14 document types, and they used it to assess mainstream VLMs with human-aligned scoring.
Chinese ancient documents, invaluable carriers of millennia of Chinese history and culture, hold rich knowledge across diverse fields but face challenges in digitization and understanding, i.e., traditional methods only scan images, while current Vision-Language Models (VLMs) struggle with their visual and linguistic complexity. Existing document benchmarks focus on English printed texts or simplified Chinese, leaving a gap for evaluating VLMs on ancient Chinese documents. To address this, we present AncientDoc, the first benchmark for Chinese ancient documents, designed to assess VLMs from OCR to knowledge reasoning. AncientDoc includes five tasks (page-level OCR, vernacular translation, reasoning-based QA, knowledge-based QA, linguistic variant QA) and covers 14 document types, over 100 books, and about 3,000 pages. Based on AncientDoc, we evaluate mainstream VLMs using multiple metrics, supplemented by a human-aligned large language model for scoring.