TY - GEN
N2 - Biometric authentication by means of handwritten signatures is a challenging pattern recognition task, which aims to infer a writer model from only a handful of genuine signatures. In order to make it more difficult for a forger to attack the verification system, a promising strategy is to combine different writer models. In this work, we propose to complement a recent structural approach to offline signature verification based on graph edit distance with a statistical approach based on metric learning with deep neural networks. On the MCYT and GPDS benchmark datasets, we demonstrate that combining the structural and statistical models leads to significant improvements in performance, profiting from their complementary properties.
DO - 10.1007/978-3-319-97785-0_45
DO - DOI
AB - Biometric authentication by means of handwritten signatures is a challenging pattern recognition task, which aims to infer a writer model from only a handful of genuine signatures. In order to make it more difficult for a forger to attack the verification system, a promising strategy is to combine different writer models. In this work, we propose to complement a recent structural approach to offline signature verification based on graph edit distance with a statistical approach based on metric learning with deep neural networks. On the MCYT and GPDS benchmark datasets, we demonstrate that combining the structural and statistical models leads to significant improvements in performance, profiting from their complementary properties.
AD - University of Fribourg, Fribourg, Switzerland
AD - University of Fribourg, Fribourg, Switzerland
AD - University of Fribourg, Fribourg, Switzerland
AD - University of Fribourg, Fribourg, Switzerland
AD - School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland
AD - University of Fribourg, Fribourg, Switzerland
AD - School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland ; DIVA group, University of Fribourg, Fribourg, Switzerland
T1 - Offline signature verification by combining graph edit distance and triplet networks
DA - 2018-08
DA - 2018-08
CY - Beijing, China
AU - Maergner, Paul
AU - Pondenkandath, Vinaychandran
AU - Alberti, Michele
AU - Liwicki, Marcus
AU - Riesen, Kaspar
AU - Ingold, Rolf
AU - Fischer, Andreas
L1 - http://hesso.tind.io/record/3233/files/Riesen_2018_offline_signature_verification_graph_edit_distance_triplets_networks.pdf
JF - Proceedings of Joint IAPR International Workshop, S+SSPR 2018, Beijing, China, 17-19 August 2018
PB - 17-19 August 2018
PP - Beijing, China
LA - eng
PY - 2018-08
PY - 2018-08
ID - 3233
L4 - http://hesso.tind.io/record/3233/files/Riesen_2018_offline_signature_verification_graph_edit_distance_triplets_networks.pdf
KW - IngĂ©nierie
KW - offline signature verification
KW - graph edit distance
KW - metric learning
KW - deep convolutional neural network
KW - triplet network
SN - 978-3-319-97784-3
TI - Offline signature verification by combining graph edit distance and triplet networks
Y1 - 2018-08
L2 - http://hesso.tind.io/record/3233/files/Riesen_2018_offline_signature_verification_graph_edit_distance_triplets_networks.pdf
LK - http://hesso.tind.io/record/3233/files/Riesen_2018_offline_signature_verification_graph_edit_distance_triplets_networks.pdf
UR - http://hesso.tind.io/record/3233/files/Riesen_2018_offline_signature_verification_graph_edit_distance_triplets_networks.pdf
ER -