Title: Ensemble of transfer learning with convolutional neural networks for writer recognition in historical documents

Authors: Radmila Janković Babić; Alessia Amelio; Ivo Rumenov Draganov; Marijana Ćosović

Addresses: Mathematical Institute of the Serbian Academy of Sciences and Arts, Kneza Mihaila 36, Belgrade City, 11000, Serbia ' HPC Laboratory, Department of Engineering and Geology, University 'G. d'Annunzio' Chieti-Pescara, Viale Pindaro 42, Pescara City, 65127, Italy ' Department of Radio Communications and Video Technologies, Technical University of Sofia, 8 Kl. Ohridski Blvd., Sofia City, 1000, Bulgaria ' Faculty of Electrical Engineering, University of East Sarajevo, Vuka Karadžića 30, Lukavica City, 71126, East Sarajevo, Bosnia and Herzegovina

Abstract: In the cultural heritage domain, writer recognition has become a challenging classification task still explored for historical documents, due to the presence of different types of noise in the documents, i.e., ink bleed-through, ink corrosion, stains on paper or parchment, difficulty in the character discrimination, elements different from the text, such as images, etc. that limit the effectiveness of existing techniques. To further advance in terms of robustness of classification and experimental setting, we propose a new deep learning model which ensembles pre-trained convolutional neural networks for writer recognition. Specifically, the ensemble is composed of three pre-trained Inception-ResNet-v2 models with different hyperparameter values. Results obtained on the benchmark ICDAR 2019 dataset of handwritten historical documents prove that the proposed approach is very promising in recognising the handwritten characters of different writers, also when compared with other deep learning models.

Keywords: convolutional neural networks; CNNs; writer recognition; cultural heritage; historical documents; ensemble learning; artificial neural networks; document analysis; deep learning; transfer learning.

DOI: 10.1504/IJRIS.2026.152162

International Journal of Reasoning-based Intelligent Systems, 2026 Vol.18 No.2, pp.86 - 100

Received: 31 Jul 2023
Accepted: 01 Jul 2024
Published online: 10 Mar 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article