Online Signature Verification: Fusion of a Hidden Markov Model and a Neural Network via a Support Vector Machine
File is not available to download!
Source: IWFHR 2002
Authors: Marc Fuentes, Sonia Garcia-Salicetti, Bernadette Dorizzi
Abstract: We describe two on-line signature verification modules. The first one considers the signatures as a sequence of points and models the true signatures of a given signer through a Hidden Markov Model (HMM). Forgeries are used to compute a decision threshold. In the second module, global parameters of the signature are the inputs of a two-classes neural network trained for each signer on both the true and "other" signatures (true signatures of other signers). Fusion of these two complementary experts through a Support Vector Machine (SVM), allows improving the results over those of each module, on a database of 51 signers.