A Comparative Study of Two State-of-the-Art Feature Selection Algorithms for Texture-Based Pixel-Labeling Task of Ancient Documents

Autor: Maroua Mehri, Ramzi Chaieb, Karim Kalti, Pierre Héroux, Rémy Mullot, Najoua Essoukri Ben Amara

Publikováno v: Journal of Imaging, Vol 4, Iss 8, p 97 (2018)

Recently, texture features have been widely used for historical document image analysis. However, few studies have focused exclusively on feature selection algorithms for historical document image analysis. Indeed, an important need has emerged to us

Externí odkaz: https://doaj.org/article/995aee0fa9bb4ce8aee2e2deff64564a

Symbols Detection and Classification using Graph Neural Networks

Autor: Muhammet Balcilar, Sébastien Adam, Paul Honeine, Pierre Héroux, Benoit Gaüzère, Guillaume Renton

Publikováno v: Pattern Recognition Letters
Pattern Recognition Letters, Elsevier, 2021, ⟨10.1016/j.patrec.2021.09.020⟩

International audience; In this paper, we propose a method to both extract and classify symbols in floorplan images. This method relies on the very recent developments of Graph Neural Networks (GNN). In the proposed approach, floorplan images are fir

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::77895fec2f0d104a1d5ec98f3f5a8f13
https://hal-normandie-univ.archives-ouvertes.fr/hal-03410511

When Spectral Domain Meets Spatial Domain in Graph Neural Networks

Autor: Muhammet, Balcilar, Guillaume, Renton, Pierre, Héroux, Benoit, Gaüzère, Sébastien, Adam, Honeine, Paul

Publikováno v: Thirty-seventh International Conference on Machine Learning (ICML 2020)-Workshop on Graph Representation Learning and Beyond (GRL+ 2020)
Thirty-seventh International Conference on Machine Learning (ICML 2020)-Workshop on Graph Representation Learning and Beyond (GRL+ 2020), Jul 2020, Vienna, Austria

International audience; Convolutional Graph Neural Networks (Con-vGNNs) are designed either in the spectral domain or in the spatial domain. In this paper, we provide a theoretical framework to analyze these neural networks, by deriving some equivale

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::c8e772ee4bb089c1e20f45dcfbe673a9
https://hal-normandie-univ.archives-ouvertes.fr/hal-03088374

New binary linear programming formulation to compute the graph edit distance

Autor: Julien Lerouge, Romain Raveaux, Sébastien Adam, Zeina Abu-Aisheh, Pierre Héroux

Publikováno v: Pattern Recognition
Pattern Recognition, Elsevier, 2017, 72, pp.254-265. ⟨10.1016/j.patcog.2017.07.029⟩

International audience; Graph edit distance (GED) is a powerful and flexible graph matching paradigm that can be used to address different tasks in structural pattern recognition, machine learning, and data mining. In this paper, some new binary line

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::065cd2dfa546135e5565f23486a0d0c2
https://doi.org/10.1016/j.patcog.2017.07.029

Graph Neural Network for Symbol Detection on Document Images

Autor: Benoit Gaüzère, Sébastien Adam, Guillaume Renton, Pierre Héroux

Publikováno v: GREC@ICDAR
13th IAPR International Workshop on Graphics Recognition
13th IAPR International Workshop on Graphics Recognition, Sep 2019, Sydney, Australia

International audience; In this paper, we propose a new method to simultaneously detect and classify symbols in floorplan images. This method relies on the very recent developments of Graph Neural Networks (GNN). In the proposed approach, floorplan i

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::40e11a5b5d59d6e698fc57e51a2e037b
https://doi.org/10.1109/icdarw.2019.00016