Image analysis – Applications – Personnel identification
Reexamination Certificate
1999-05-26
2001-11-13
Johns, Andrew W. (Department: 2621)
Image analysis
Applications
Personnel identification
C382S253000
Reexamination Certificate
active
06317507
ABSTRACT:
The invention relates to a device for the verification of time-dependent, user-specific signals which includes
means for generating a set of feature vectors which serve to provide an approximative description of an input signal and are associated with selectable sampling intervals of the signal,
means for preparing a hidden Markov model (HMM) for the signal,
means for determining a probability value which describes the probability of occurrence of the set of feature vectors, given the HMM, and
a threshold decider for comparing the probability value with a threshold value and for deciding on the verification of the signal.
For the verification of time-dependent, user-specific signals, notably signatures or speech signals, it is checked whether an input signal indeed originates from a specific user or is a forgery. In this context the term “time dependency” is to be understood to mean that the signals are generated by the user in a given time interval, specific, different signal components being associated with different instants within the time interval. Before verification can take place, a signal model must be formed by means of one or more original signals; for this purpose use is made of so-called hidden Markov models (HMMs). The original signals used for forming the model are training signals for the so-called training of the HMM model. After completion of training, a signal can be verified by means of the device. To this end, a user identification, for example a user name or a number assigned to a user, is entered on the one hand and the user-specific signal on the other hand. The input signal is transformed into a set of feature vectors. In order to form the vector components in the case of signatures, for example co-ordinates passed during the writing of the signature are evaluated and also the pressure exerted by an input stylus. Subsequently, there is formed a probability value which describes the probability of occurrence of the set of feature vectors for the HMM model assigned to the user with the user identification. The input signal is recognized as an original signal up to a selectable threshold value, and beyond that as a forgery.
It is an object of the invention to construct a device of the kind set forth in such a manner that the verification is better adapted to the relevant user and to the user-specific signals assigned to said user.
This object is achieved in that means are provided for the automatic, person-specific evaluation of the relevance of the features collected in the feature vectors by means of an LDA transformation, and for the automatic selection of features evaluated as relevant.
The so-called LDA (Linear Discriminant Analysis) transformation is a known transformation which is described in, for example K. Fukunaga “Introduction to Statistical Pattern Recognition”, Second Edition, Academic Press, New York, 1990, chapter 10.2. The transformed feature vectors formed on the basis thereof have components which are arranged in conformity with their relevance to the characterization of the respective user-specific signal. A selectable number of components of the feature vectors can thus be selected in a manner adapted to the relevant circumstances, resulting in feature vectors of reduced dimension. Neither forgeries of the user-specific signal nor user-specific signals of other users are taken into account for the selection. The reduction of the feature vectors by way of the LDA transformation is accompanied by a reduction of the HMM model used; this leads to a faster verification operation and a reduction of the required storage capacity. Generally speaking, the term probability value is to be understood to describe a value which is derived from a probability, notably the original value of the probability, or a logarithmic value of the probability.
The invention is preferably used for on-line verification, but is also suitable for off-line verification. The user-specific signals are, for example, signatures or speech signals.
REFERENCES:
patent: 5839103 (1998-11-01), Mammone et al.
patent: 5848388 (1998-12-01), Power et al.
patent: 5995927 (1999-11-01), Li
patent: 0744734A2 (1996-11-01), None
A Two-Stage Procedure for Phone Based Speaker Verification, by Jesper Olsen, pp. 889-897; Pattern Recognition Letters, Bd. 18, Nr. 9, Sep. 1, 1997.
Netsch L.P. et al, “Speaker Verification Using Temporal Decorrelation Post-Processing”, Database Accession No. 4441422; XP-002121311.
“Application of Hidden Markov-Models for Signature Verification”, L. Yang, B. Widjaja and R. Prasad, Pattern Recognition 28, pp. 161-170.
“Introduction to Statistical Pattern Recognition”, Second Edition, Academic Press, New York, 1990, Chapter 10.2.
“Fundamentals of Speech Recognition”, First Edition, Prentice Hall, 1993, L.R. Rabiner and B.H. Juang, Chapters 6.4-6.6.
Azarian Seyed
Johns Andrew W.
Piotrowski Daniel J.
U.S. Philips Corporation
LandOfFree
Signal verification device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Signal verification device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Signal verification device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2613034