An archaeological Catalog Collection Method Based on Large Vision-Language Models
This work addresses the challenge of automating catalog collection for archaeologists, but it appears incremental as it builds on existing VLMs with specific modules for archaeological data.
The paper tackles the problem of automated collection of archaeological catalogs, which are scattered across publications, by proposing a method based on Large Vision-Language Models with modules for document localization, block comprehension, and block matching, achieving effectiveness as demonstrated through data collection from Dabagou and Miaozigou pottery catalogs and comparison experiments.
Archaeological catalogs, containing key elements such as artifact images, morphological descriptions, and excavation information, are essential for studying artifact evolution and cultural inheritance. These data are widely scattered across publications, requiring automated collection methods. However, existing Large Vision-Language Models (VLMs) and their derivative data collection methods face challenges in accurate image detection and modal matching when processing archaeological catalogs, making automated collection difficult. To address these issues, we propose a novel archaeological catalog collection method based on Large Vision-Language Models that follows an approach comprising three modules: document localization, block comprehension and block matching. Through practical data collection from the Dabagou and Miaozigou pottery catalogs and comparison experiments, we demonstrate the effectiveness of our approach, providing a reliable solution for automated collection of archaeological catalogs.