Update strategies for HMM-based dynamic signature biometric systems

Abstract

Esta es la versión de autor de la comunicación de congreso publicada en: This is an author produced version of a paper published in: Abstract-Biometric authentication on devices such as smartphones and tablets has increased significantly in the last years. One of the most acceptable and increasing traits is the handwriting signature as it has been used in financial and legal agreements scenarios for over a century. Nowadays, it is frequent to sign in banking and commercial areas on digitizing tablets. For these reasons, it is necessary to consider a new scenario where the number of training signatures available to generate the user template is variable and besides it has to be taken into account the lap of time between them (inter-session variability). In this work we focus on dynamic signature verification. The main goal of this work is to study system configuration update strategies of time functions-based systems such as Hidden Markov Model (HMM) and Gaussian Mixture Models (GMM). Therefore, two different cases have been considered. First, the usual case of having an HMM-based system with a fixed configuration (i.e. Baseline System). Second, an HMM-based and GMM-based systems whose configurations are optimized regarding the number of training signatures available to generate the user template. The experimental work has been carried out using an extended version of the Signature Long-Term database taking into account skilled and random or zero-effort forgeries. This database is comprised of a total of 6 different sessions distributed in a 15-month time span. Analyzing the results, the Proposed Systems achieve an average absolute improvement of 4.6% in terms of EER(%) for skilled forgeries cases compared to the Baseline System whereas the average absolute improvement for the random forgeries cases is of 2.7% EER. These results show the importance of optimizing the configuration of the systems compared to a fixed configuration system when the number of training signatures available to generate the user template increases.

Automatic Signature Verification in the Mobile Cloud Scenario: Survey and Way Ahead

“…However results obtained by the pattern recognition community within the re-training process (in semi-supervised scenarios) haven't been studied within this specific field, so that it must be considered to be a challenging open issue [47]. Some results can be referred to HMM and GMM [92], (even if in this case signature where acquired only by means of a single platform): the standard case of having an HMM-based system with a fixed configuration and an HMMbased and a GMM-based system with optimized configurations in function of the training signatures available at enrolment stage. The approach has been able to let the system achieves an average absolute improvement of 4.6% (2.7%) in terms of EER, with respect to the baseline system, for the skilled (random) forgery cases.…”

Reducing the template ageing effect in on‐line signature biometrics

“…Three well-known systems based on previous studies are considered here: HMM, GMM, and DTW. In all of them, signals captured by the digitiser (only X and Y coordinates and pressure) are used to extract a set of 23 time functions for each signature [14]. Time functions related to pen angular orientation (azimuth and altitude angles) were discarded in order to consider the same set of time functions that we would be able to use in general purpose devices such as tablets and smartphones.…”

“…• This paper reports the first significant experimental results regarding the effect of template ageing and template update strategies for on-line signature authentication considering both random and skilled forgeries of the signatures. For this, we have created an extension of the ATVS On-Line Signature Long-Term database in which skilled forgeries are included [14]. The complete signature database is publicly available at https:// github.com/BiDAlab/xLongSignDB.…”

Complexity-Based Biometric Signature Verification

“…For each signature acquired using a digitizing tablet (see Sec. III), signals related to X and Y pen coordinates and pressure are used to extract a set of 23 time functions, similar to [16] (see Table I). The more discriminative and robust time functions of each complexity level are selected using the Sequential Forward Feature Selection algorithm (SFFS) enhancing the signature verification system in terms of EER.…”