Confidence voting method ensemble applied to off-line signature verification

Abstract

In this paper, a new approximation to off-line signature verification is proposed based on two-class classifiers using an expert decisions ensemble. Different methods to extract sets of local and a global features from the target sample are detailed. Also a normalization by confidence voting method is used in order to decrease the final equal error rate (EER). Each set of features is processed by a single expert, and on the other approach proposed, the decisions of the individual classifiers are combined using weighted votes. Experimental results are given using a subcorpus of the large MCYT signature database for random and skilled forgeries. The results show that the weighted combination outperforms the individual classifiers significantly. The best EER obtained were 6.3 % in the case of skilled forgeries and 2.31 % in the case of random forgeries.

Automatic Online Signature Verification Based Only on FHE Features: An Oxymoron?

“…Other works proposed different information fusion strategies. For instance, the sum and max rule are proposed in [17] for online signature verification, while a confidence voting method is proposed in [18] for the offline case. In this paper, two different verification schemes are proposed.…”

Static Signature Verification Based on Texture Analysis Using Support Vector Machine

“…which uses digitizing tablet to collect signature information. Based on the knowledge of the forger about the signature and signer, forgery can be broadly classified into three types such as: random, simple and skilled (Diaz-Cabrera, Morales, & Ferrer, 2014); Rico-Juan & Iñesta, 2012). The forgeries are posing a major threat for certification and validation of a person through signature.…”

Improving classification using a Confidence Matrix based on weak classifiers applied to OCR

“…Each weak classifier is trained using a leaving-one-out scheme: each single sample is isolated from the training set T and the rest are used in combination with the NN to produce the confidence values. The formula detailed below is used in Rico-Juan and Iñesta [23] and is inspired by Pérez-Cortés et al [20]. If x is a training sample, then the confidence value for each class w ∈ Ω is based on the following equation:…”