Publication Details
Importance of Textlines in Historical Document Classification
Kohút Jan, Ing. (DCGM FIT BUT)
Beneš Karel, Ing. (DCGM FIT BUT)
Hradiš Michal, Ing., Ph.D. (DCGM FIT BUT)
Historical document classification, Script and font classification, Document origin localization, Document dating.
This paper describes a system prepared at Brno University of Technology for ICDAR 2021 Competition on Historical Document Classification, experiments leading to its design, and the main findings. The solved tasks include script and font classification, document origin localization, and dating. We combined patch-level and line-level approaches, where the line-level system utilizes an existing, publicly available page layout analysis engine. In both systems, neural networks provide local predictions which are combined into page-level decisions, and the results of both systems are fused using linear or log-linear interpolation. We propose loss functions suitable for weakly supervised classification problem where multiple possible labels are provided, and we propose loss functions suitable for interval regression in the dating task. The line-level system significantly improves results in script and font classification and in the dating task. The full system achieved 98.48%, 88.84%, and 79.69% accuracy in the font, script, and location classification tasks respectively. In the dating task, our system achieved a mean absolute error of 21.91 years. Our system achieved the best results in all tasks and became the overall winner of the competition.
@INPROCEEDINGS{FITPUB12706, author = "Martin Ki\v{s}\v{s} and Jan Koh\'{u}t and Karel Bene\v{s} and Michal Hradi\v{s}", title = "Importance of Textlines in Historical Document Classification", pages = "158--170", booktitle = "Uchida, S., Barney, E., Eglin, V. (eds) Document Analysis Systems", series = "Lecture Notes in Computer Science", volume = 13237, year = 2022, location = "La Rochelle, FR", publisher = "Springer Nature Switzerland AG", ISBN = "978-3-031-06554-5", doi = "10.1007/978-3-031-06555-2\_11", language = "english", url = "https://www.fit.vut.cz/research/publication/12706" }