A Cross-Font Image Retrieval Network for Recognizing Undeciphered Oracle Bone Inscriptions

2024-09-10

arXiv.org

AI Summary

The large number of undeciphered oracle bone inscriptions (OBIs) poses a significant challenge to efficient and accurate paleographic interpretation. Method: This paper proposes a cross-glyph image retrieval method tailored for ancient character decipherment. We design a Siamese network architecture that jointly integrates Multi-scale Feature Integration (MFI) and Multi-scale Refined Classification (MRC) to model structural invariance across resolutions, enabling visual-semantic alignment between undeciphered OBIs and well-established ancient scripts (e.g., bronze inscriptions and small seal script). Contribution/Results: Our work introduces the first “cross-glyph retrieval” paradigm—balancing interpretability and reusability. Evaluated on three cross-glyph retrieval benchmarks, the method significantly outperforms existing baselines, achieving precise matching between undeciphered and known characters. It delivers the first end-to-end, verifiable AI-assisted tool for OBI interpretation, supporting reproducible and traceable paleographic analysis.

Technology Category

• Computer Vision: Image and Video Retrieval
• Reasoning under Uncertainty: Relational Probabilistic Models
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Oracle Bone Inscription (OBI) is the earliest mature writing system in China, which represents a crucial stage in the development of hieroglyphs. Nevertheless, the substantial quantity of undeciphered OBI characters remains a significant challenge for scholars, while conventional methods of ancient script research are both time-consuming and labor-intensive. In this paper, we propose a cross-font image retrieval network (CFIRN) to decipher OBI characters by establishing associations between OBI characters and other script forms, simulating the interpretive behavior of paleography scholars. Concretely, our network employs a siamese framework to extract deep features from character images of various fonts, fully exploring structure clues with different resolutions by multiscale feature integration (MFI) module and multiscale refinement classifier (MRC). Extensive experiments on three challenging cross-font image retrieval datasets demonstrate that, given undeciphered OBI characters, our CFIRN can effectively achieve accurate matches with characters from other gallery fonts, thereby facilitating the deciphering.

Problem

• oracle bone script recognition
• ancient Chinese writing
• time-consuming methods

Innovation

• CFIRN
• Oracle Bone Script Recognition
• Cross-Font Image Retrieval