Recurrent binary patterns and CNNs for offline signature verification
| dc.contributor.author | Yılmaz, Mustafa Berkay | |
| dc.contributor.author | Özturk, Kağan | |
| dc.date.accessioned | 2021-02-19T21:16:33Z | |
| dc.date.available | 2021-02-19T21:16:33Z | |
| dc.date.issued | 2020 | |
| dc.department | ALKÜ | |
| dc.description | Future Technologies Conference (FTC) -- OCT 24-25, 2019 -- San Francisco, CA | |
| dc.description | Yilmaz, Mustafa Berkay/0000-0003-0320-3957 | |
| dc.description.abstract | Signature representations that are extracted by convolutional neural networks (CNN) can achieve low error rates. However, a trade-off exists between such models' complexities and hand-crafted features' slightly higher error rates. A novel writer-dependent (WD) recurrent binary pattern (RBP) network, and a novel signer identification CNN is proposed. RBP network is a recurrent neural network (RNN) to learn the sequential relation between binary pattern histograms over image windows. A novel histogram selection method is introduced to remove the stop-word codes. Dimensionality is reduced by more than 25% while improving the results. This work is the first to combine binary patterns and RNNs for static signature verification. Several test sets, derived from large-scale and popular databases (GPDS-960 and GPDS-Synthetic-10000) are used. Without training any global classifier, RBP network provides competitive equal error rates (EER). The proposed architectures are compared and integrated with other recent CNN models. Score-level integration of WD classifiers trained with different representations are investigated. Cross-validation tests demonstrate the EERs reduced compared to the best single classifier. A state-of-the-art EER of 1.11% is reported with a global decision threshold (0.57% EER with user-based thresholds) on GPDS-160 database. | |
| dc.description.sponsorship | Scientific Research Projects Coordination Unit of Akdeniz UniversityAkdeniz University [3780] | |
| dc.description.sponsorship | This work was supported by The Scientific Research Projects Coordination Unit of Akdeniz University, project number: 3780. | |
| dc.identifier.doi | 10.1007/978-3-030-32523-7_29 | |
| dc.identifier.endpage | 434 | en_US |
| dc.identifier.isbn | 978-3-030-32523-7; 978-3-030-32522-0 | |
| dc.identifier.issn | 2194-5357 | |
| dc.identifier.issn | 2194-5365 | |
| dc.identifier.scopusquality | N/A | |
| dc.identifier.startpage | 417 | en_US |
| dc.identifier.uri | https://doi.org/10.1007/978-3-030-32523-7_29 | |
| dc.identifier.uri | https://hdl.handle.net/20.500.12868/472 | |
| dc.identifier.volume | 1070 | en_US |
| dc.identifier.wos | WOS:000560948600029 | |
| dc.identifier.wosquality | N/A | |
| dc.indekslendigikaynak | Web of Science | |
| dc.indekslendigikaynak | Scopus | |
| dc.institutionauthor | 0-belirlenecek | |
| dc.language.iso | en | |
| dc.publisher | Springer International Publishing Ag | |
| dc.relation.ispartof | Proceedings of the Future Technologies Conference (Ftc) 2019, Vol 2 | |
| dc.relation.ispartofseries | Advances in Intelligent Systems and Computing | |
| dc.relation.publicationcategory | Konferans Öğesi - Uluslararası - Kurum Öğretim Elemanı | |
| dc.rights | info:eu-repo/semantics/closedAccess | |
| dc.subject | Offline signature verification | |
| dc.subject | Recurrent neural network | |
| dc.subject | Convolutional neural network | |
| dc.subject | Score-level integration | |
| dc.title | Recurrent binary patterns and CNNs for offline signature verification | |
| dc.type | Conference Object |
