Signature recognition

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

Approach of face recognition based on continuous hidden Markov model

Abstract: A human face recognition approach based on the continuous hidden Markov model is proposed,its essential contents list as follows:①With the stability and the transposes the invariability,singular value decomposition is used to extract the character of the face image as the observed value array for the normalized face image.②In the face automatic recognition system,a continuous HMM method is used to recognize face and the two Gaussian pdf is adopted to build up the HMM model,and then the model could be used in identifying face.The results of experiments show that the method proposed reduces the computing time,runs faster,and the recognition rate is in-creased,so it can satisfy the needs of real-time face recognition.

Real-time automatic speech recognition using HMM and neural networks

Abstract: The authors describe the implementation of an automatic speech recognition system for isolated words, based on the integration of neural networks (NN) for phonetic mapping with hidden Markov models (HMMs) for word modeling. The main system features and implementation algorithms are presented, and initial experimental results and software capabilities are shown. It is shown that the inclusion of the NN model between the acoustic processor and the HMM improves the recognition and avoids the clustering and labelling phases. >

Template-based and HMM-based Approaches for Isolated Spanish Digit Recognition

Abstract: Isolated word recognition is usually performed by two dieren t approaches, viz., Conventional Templatebased and Hidden Markov Models (HMM)-based. A comparison between these approaches for isolated digit recognition (in Spanish language) is performed in this paper. For the template-based approach, several Dynamic Time Warping algorithms are proposed and implemented in a Matlab environment, while for the HMM-based approach, the algorithms available in a freeware toolbox has been employed. The algorithms are tested on a database collected during a Course on Digital Processing of Speech Signals at the University of Rosario. The test results show that a better performance can be obtained with a HMM-based digit recognizer.

Off-Line Writer Verification Using Segments of Handwritten Samples and SVM

Abstract: This works presents a method to verify a person identity based on off-line handwritten strokes analysis. Its main contribution is that the descriptors are obtained from the constitutive segments of each grapheme, in contrast with the complexity of the handwritten images used in signature recognition system or even with the graphemes themselves. In this way, only few handwriting samples taken from a short text could be enough to identify the writer. The descriptor is based on an estimation of the pressure of the stroke grayscale image. In particular, the average of the gray levels on the perpendicular line to the skeleton is used. A semi-automatic procedure is used to extract the segments from scanned images. The repository consists of 3.000 images of 6 different segments. Binary-output Support Vector Machine classifiers are used. Two types of cross validation, K-fold and Leave-one-out, are implemented to objectively evaluate the descriptor performance. The results are encouraging. A hit rate of 98% in identity verification is obtained for the 6 segments studied.