Signature recognition

About: Signature recognition is a research topic. Over the lifetime, 2138 publications have been published within this topic receiving 37605 citations.

Review on Offline Signature Recognition and Verification Techniques

Abstract: 1. D. Bertolini, L. S. Oliveira, E. Justino, and R. Sabourin. Reducing forgeries in writer-independent off-line signature verification through ensemble of classifiers. Pattern Recognition, 43(1), January 2010. 2. R.K. Bharathi and B.H. Shekar. Off-line signature verification based on chain code histogram and Support Vector Machine. In 2013 International Conference on Advances in Computing, Communications and Informatics (ICACCI), pages 2063–2068, August 2013 3. Alceu S. Britto, Robert Sabourin, and Luiz E. S. Oliveira. Dynamic selection of classifiers -

Hybrid Wavelets based Feature Vector Generation from Multidimensional Data set for On-line Handwritten Signature Recognition

Abstract: On-line handwritten Signature is one of the important behavioural biometric trait. On-line signature have more information such as x, y, z variations, pressure levels, Azimuth and Altitude of pen tip, due to this better accuracy can be achieved when signatures are captured in real time with digitizer device. In this paper a technique based on Hybrid Wavelets to extract texture features of Dynamic Handwritten (On-line) signature is proposed. Hybrid wavelets are flexible and combine the advantage of transforms and Multiresolution analysis. Proposed system uses the hybrid wavelets to generate the wavelet energy distribution of the pressure pattern of dynamic signatures, velocity magnitude, Azimuth & Altitude vectors. Hybrid Wavelet of Type I and Type II are used and their performance is compared. Hybrid Wavelets are found to give highest Performance Index of 83.96% for Azimuth and Altitude based feature vector.

Hidden Model Sequence Models for Automatic Speech Recognition

Abstract: Most modern automatic speech recognition systems make use of acoustic models based on hidden Markov models. To obtain reasonable recognition performance within a large vocabulary framework, the acoustic models usually include a pronunciation model, together with complex parameter tying schemes. In many cases the pronunciation model operates on a phoneme level and is derived independently of the underlying models. In contrast, this work is aimed at improving pronunciation modelling on a sub-phone level in a combined framework. The modelling of pronunciation variation is assumed to be of special importance for recognition of spontaneous speech. The novel framework of hidden model sequence modelling using the basis of hidden Markov models is presented. The framework allows arbitrary mappings between model and phoneme sequences. A stochastic model for this mapping, the model sequence model, is introduced. The parameters of this new model can be estimated jointly with the parameters of the underlying hidden Markov model set by the use of a maximum likelihood criterion. The set of potential mappings is divided into the cases of fixed and variable alignment between the model and phoneme sequences. Whereas the former only allows the modelling of substitutions, the latter can be used to describe additional sub-phone insertion and deletion effects. A range of different approaches for potential model sequence models is proposed. The natural form of modelling the fixed alignment case is an N-gram type model using a constrained phoneme context. Issues such as model structure, data sparsity and appropriate initialisation of the model parameters are discussed in detail and a set of solutions is tested. For the case of variable alignment a multigram based model sequence model is presented and two different implementations are investigated. Special attention is given to the modelling of sub-phone deletions. Experimental evidence is presented on the basis of results obtained on two transcription tasks: Resource Management as an example for read speech; and Switchboard as an example for a complex spontaneous speech task. On both tasks statistically significant improvements are obtained over a standard HMM baseline with a relative reduction in word error rate of more than 25% on Resource Management and 5% on Switchboard.

Biometric Security Systems: Fingerprint and Speech Technology.

Abstract: This paper deals with the design of a biometric security system based upon the fingerprint and speech technology. In the first chapter there are the biometric security systems and a concept of an integration of the both technologies introduced. Then the fingerprint technology followed by the speech technology is shortly described. There are discussed some basic principles of each of the technologies.

Automatic speaker recognition by computers

Abstract: Summary form only given, as follows. This paper describes a new modular speaker recognition system consisting of a st of real?time speech analysis processors and a pattern recognition software package. Within a government sponsored research project, combinations of different speech analysis procedures and different pattern recognition algorithms are compared in order to find optimal subsystems, to be applied to security systems or law enforcement, for given boundary conditions. In order to find the influence of different techniques, distance measures, quantisation band distortions on the recognition rate of given data base (2,500 utterances), a study has been carried out, results of which are being discussed.