Fully convolutional neural networks for newspaper article segmentation

Meier, Benjamin; Stadelmann, Thilo; Stampfli, Jan; Arnold, Marek; Cieliebak, Mark

doi:10.21256/zhaw-1533

Please use this identifier to cite or link to this item: https://doi.org/10.21256/zhaw-1533

Publication type:	Conference paper
Type of review:	Peer review (publication)
Title:	Fully convolutional neural networks for newspaper article segmentation
Authors:	Meier, Benjamin Stadelmann, Thilo Stampfli, Jan Arnold, Marek Cieliebak, Mark
DOI:	10.21256/zhaw-1533
Proceedings:	Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)
Conference details:	14th IAPR International Conference on Document Analysis and Recognition (ICDAR 2017), Kyoto Japan, 13-15 November 2017
Issue Date:	2017
Publisher / Ed. Institution:	CPS
Publisher / Ed. Institution:	Kyoto
Language:	English
Subjects:	Semantic segmentation; CNN; Deep learning; Datalab
Subject (DDC):	006: Special computer methods
Abstract:	Segmenting newspaper pages into articles that semantically belong together is a necessary prerequisite for article-based information retrieval on print media collections like e.g. archives and libraries. It is challenging due to vastly differing layouts of papers, various content types and different languages, but commercially very relevant for e.g. media monitoring. We present a semantic segmentation approach based on the visual appearance of each page. We apply a fully convolutional neural network (FCN) that we train in an end-to-end fashion to transform the input image into a segmentation mask in one pass. We show experimentally that the FCN performs very well: it outperforms a deep learning-based commercial solution by a large margin in terms of segmentation quality while in addition being computationally two orders of magnitude more efficient.
URI:	https://digitalcollection.zhaw.ch/handle/11475/1863
Fulltext version:	Published version
License (according to publishing contract):	Licence according to publishing contract
Departement:	School of Engineering
Organisational Unit:	Institute of Computer Science (InIT)
Appears in collections:	Publikationen School of Engineering

Files in This Item:

File	Description	Size	Format
212962.pdf		523.17 kB	Adobe PDF	View/Open

Show full item record

Meier, B., Stadelmann, T., Stampfli, J., Arnold, M., & Cieliebak, M. (2017). Fully convolutional neural networks for newspaper article segmentation. Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR). https://doi.org/10.21256/zhaw-1533

Meier, B. et al. (2017) ‘Fully convolutional neural networks for newspaper article segmentation’, in Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR). Kyoto: CPS. Available at: https://doi.org/10.21256/zhaw-1533.

B. Meier, T. Stadelmann, J. Stampfli, M. Arnold, and M. Cieliebak, “Fully convolutional neural networks for newspaper article segmentation,” in Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR), 2017. doi: 10.21256/zhaw-1533.

MEIER, Benjamin, Thilo STADELMANN, Jan STAMPFLI, Marek ARNOLD und Mark CIELIEBAK, 2017. Fully convolutional neural networks for newspaper article segmentation. In: Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR). Conference paper. Kyoto: CPS. 2017

Meier, Benjamin, Thilo Stadelmann, Jan Stampfli, Marek Arnold, and Mark Cieliebak. 2017. “Fully Convolutional Neural Networks for Newspaper Article Segmentation.” Conference paper. In Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR). Kyoto: CPS. https://doi.org/10.21256/zhaw-1533.

Meier, Benjamin, et al. “Fully Convolutional Neural Networks for Newspaper Article Segmentation.” Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR), CPS, 2017, https://doi.org/10.21256/zhaw-1533.