Signature recognition

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

Automatic speech recognition using artificial life

Abstract: After years of research activity, the machine recognition performance of speech still does not match human performance. As speech recognition is a pattern recognition problem, classification is an important part of any speech recognition system. This paper, briefly reviews a number of classification methods that have been used in automatic speech recognition systems and proposes a new back-end classifier that is based on artificial life. The paper describes how the proposed classifier can be used in a speech recognition system.

Physiologically-Motivated Feature Extraction Methods for Speaker Recognition

Abstract: PHYSIOLOGICALLY-MOTIVATED FEATURE EXTRACTION METHODS FOR SPEAKER RECOGNITION Jianglin Wang, B.S., M.S. Marquette University, 2013 Speaker recognition has received a great deal of attention from the speech community, and significant gains in robustness and accuracy have been obtained over the past decade. However, the features used for identification are still primarily representations of overall spectral characteristics, and thus the models are primarily phonetic in nature, differentiating speakers based on overall pronunciation patterns. This creates difficulties in terms of the amount of enrollment data and complexity of the models required to cover the phonetic space, especially in tasks such as identification where enrollment and testing data may not have similar phonetic coverage. This dissertation introduces new features based on vocal source characteristics intended to capture physiological information related to the laryngeal excitation energy of a speaker. These features, including RPCC, GLFCC and TPCC, represent the unique characteristics of speech production not represented in current state-of-the-art speaker identification systems. The proposed features are evaluated through three experimental paradigms including cross-lingual speaker identification, cross song-type avian speaker identification and mono-lingual speaker identification. The experimental results show that the proposed features provide information about speaker characteristics that is significantly different in nature from the phonetically-focused information present in traditional spectral features. The incorporation of the proposed glottal source features offers significant overall improvement to the robustness and accuracy of speaker identification tasks.

Computerized signature verification system

Abstract: The fine structure of the human muscle forces that are exerted during the writing of a signature is consistent and well defined for most people. On the basis of this observation, an experimental system that utilizes a person's signature waveform dynamics for identification is proposed. The system is intended to be used for online signature verification. It can successfully verify a person's identity and can also detect forgeries. The acceptance rate for random forgeries, i.e. accidental matching of two different persons' signatures, is very low. >

Appearance-Based Biometric Recognition: Secure Authentication and Cancellability

Abstract: Although, biometrics provide high-confidence and trusted security, they suffer from a fatal weakness that emerges from permanence and limitation in quantities. Such a drawback puts biometric data under a substantial risk of fraudulent, which makes the replacement of traditional authentication systems infeasible with the lack of proper biometric data protection. This paper presents a novel biometric protection method to generate secure facial biometric templates used in statistical-based recognition algorithms such as 2DPCA. Original biometrics are polynomially transformed to the secure domain, where cooccurrence matrices are used to generate the final templates. The paper presents a unique relationship established by the Hadamard product within the transformation. The generated secure templates are used in the same fashion as original biometrics for evaluations using 2DPCA without any change to the recognition algorithm. And yet, evaluations confirm high security with enhanced recognition accuracy by 3% and 4.5% over the original and other transformed data respectively.

A comparative study of feature level fusion strategies for Multimodal Biometric Systems based on Face and Iris

Abstract: With the technology advances, new approaches for automatic recognition of a person's identity have been proposed and such a fact has encouraged the use of Biometrics Systems. This approach uses physical or behavioural characteristics of the user in order to recognize or authenticate their identity. The Biometric Systems can be classified as Unimodal or Multimodal. The Unimodal Systems use a single biometric modality to perform the recognition, while the Multimodal ones use two or more modalities. A Multimodal Biometric System can be constructed in different ways, according to its architecture, fusion level and fusion strategies. The main of this work is to investigate and compare different feature level fusion strategies, in order to design a Multimodal Biometric System with high performance. In this paper, we used the discrete wavelet transform to extract the feature sets from iris and face images. Experimental results show that Multimodal Biometric Systems outperform Unimodal Biometric Systems according to recognition rate computed over the outputs produced by the induced Support Vector Machine classifier.