Leveraging Morphology for Historical Script Metrological Analysis
2026-06-08 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors present a new method for analyzing historical handwriting by learning character shapes from whole lines of text rather than individual characters. They use a transformer-based model to detect and reconstruct characters, which helps measure how characters and their parts appear and vary across documents. Their approach requires only minimal training data and can distinguish subtle writing differences, useful for studying old manuscripts. They tested this on a 14th-century document and showed it can track handwriting styles efficiently.
handwritten text recognitionpaleographytransformer architecturecharacter prototypesline-level transcriptionscript morphologybounding box predictionbi-gramshistorical manuscriptsdeep learning
Authors
Malamatenia Vlachou Efstathiou, Raphaël Baena, Dominique Stutzmann, Mathieu Aubry
Abstract
Advances in handwritten text recognition have enabled large-scale transcription of historical documents, but still provide limited access to interpretable visual measurements for paleography, the study of historical scripts. In this paper, our main insight is that morphological script analysis, in particular the capacity to learn character prototypes from line-level transcriptions, enables the definition of scalable, meaningful, and stable paleographic measurements. More precisely, we leverage a transformer-based detection architecture together with a prototype-based line reconstruction module to learn prototypical characters and their occurrence, deformation, and positioning. Our contributions are twofold. First, we introduce a deep architecture and learning methodology that enables efficient character modeling with only line-level transcription supervision, significantly improving over the Learnable Typewriter baseline and enabling accurate character bounding box prediction, unlocking its potential for paleographic measurements. Second, we introduce and demonstrate the paleographical relevance of automatic measurements enabled by our architecture for characters, bi-grams, and spaces between graphical units. For this demonstration, we extend the annotations of the codex Paris, BnF, fr. 2813, commissioned in the late fourteenth century by Charles V and copied by four hands, to 160 pages. We visualize our measurements over these pages, showing how they enable us not only to differentiate graphical profiles, but also to discover and analyze subtle variations. This case study outlines the scalability of our approach and its frugality in terms of required training data, since a single column of text is sufficient to compute our measurements on each of the 160 pages. Data and code are publicly available at: https://malamatenia.github.io/morphology4metrology-analysis.