Dissimilarity representations allow for building good classifiers
Pattern Recognition Letters
An Off-Line Signature Verification System using Hidden Markov Model and Cross-Validation
SIBGRAPI '00 Proceedings of the 13th Brazilian Symposium on Computer Graphics and Image Processing
Inkteractors: interacting with digital ink
Proceedings of the 2008 ACM symposium on Applied computing
Hi-index | 0.01 |
There are different methods for signature verification proposed in the literature. Most of them, take into account a personal model, i.e., they need a considerable number of genuine signatures of the same writer to correctly train the model. This is the main drawback of this kind of approach, since in real applications we have small number of samples available for training. In this paper we propose an off-line signature verification method based on Forensic Questioned Document Examination approach. This kind of strategy reduces any classification problem to a 2-class problem, hence, makes it possible to build robust signature verification systems even when few signatures per writer are available. Comprehensive results on a database composed of 240 writers (40 samples per writer) demonstrate the efficiency of the proposed method.