J.A. du Preez

Offline signature verification using the discrete radon transform and a hidden Markov model

We developed a system that automatically authenticates offline handwritten signatures using the discrete Radon transform (DRT) and a hidden Markov model (HMM). Given the robustness of our algorithm and the fact that only global features are considered, satisfactory results are obtained. Using a database of 924 signatures from 22 writers, our system achieves an equal error rate (EER) of 18% when only high-quality forgeries (skilled forgeries) are considered and an EER of 4.5% in the case of only casual forgeries. These signatures were originally captured offline. Using another database of 4800 signatures from 51 writers, our system achieves an EER of 12.2% when only skilled forgeries are considered. These signatures were originally captured online and then digitally converted into static signature images. These results compare well with the results of other algorithms that consider only global features.

Estimating the pen trajectories of static signatures using hidden Markov models

Static signatures originate as handwritten images on documents and by definition do not contain any dynamic information. This lack of information makes static signature verification systems significantly less reliable than their dynamic counterparts. This study involves extracting dynamic information from static images, specifically the pen trajectory while the signature was created. We assume that a dynamic version of the static image is available (typically obtained during an earlier registration process). We then derive a hidden Markov model from the static image and match it to the dynamic version of the image. This match results in the estimated pen trajectory of the static image.

Efficient training of high-order hidden Markov models using first-order representations

Abstract We detail an algorithm (ORED) that transforms any higher-order hidden Markov model (HMM) to an equivalent first-order HMM. This makes it possible to process higher-order HMMs with standard techniques applicable to first-order models. Based on this equivalence, a fast incremental algorithm (FIT) is developed for training higher-order HMMs from lower-order models, thereby avoiding the training of redundant parameters. We also show that the FIT algorithm results in much faster training and better generalization compared to conventional high-order HMM approaches. This makes training of high-order HMMs practical for many applications.

Efficient backward decoding of high-order hidden Markov models

The forward-backward search (FBS) algorithm [S. Austin, R. Schwartz, P. Placeway, The forward-backward search algorithm, in: Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, 1991, pp. 697-700] has resulted in increases in speed of up to 40 in expensive time-synchronous beam searches in hidden Markov model (HMM) based speech recognition [R. Schwartz, S. Austin, Efficient, high-performance algorithms for N-best search, in: Proceedings of the Workshop on Speech and Natural Language, 1990, pp. 6-11; L. Nguyen, R. Schwartz, F. Kubala, P. Placeway, Search algorithms for software-only real-time recognition with very large vocabularies, in: Proceedings of the Workshop on Human Language Technology, 1993, pp. 91-95; A. Sixtus, S. Ortmanns, High-quality word graphs using forward-backward pruning, in: Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, 1999, pp. 593-596]. This is typically achieved by using a simplified forward search to decrease computation in the following detailed backward search. FBS implicitly assumes that forward and backward searches of HMMs are computationally equivalent. In this paper we present experimental results, obtained on the CallFriend database, that show that this assumption is incorrect for conventional high-order HMMs. Therefore, any improvement in computational efficiency that is gained by using conventional low-order HMMs in the simplified backward search of FBS is lost. This problem is solved by presenting a new definition of HMMs termed a right-context HMM, which is equivalent to conventional HMMs. We show that the computational expense of backward Viterbi-beam decoding right-context HMMs is similar to that of forward decoding conventional HMMs. Though not the subject of this paper, this allows us to more efficiently decode high-order HMMs, by capitalising on the improvements in computational efficiency that is obtained by using the FBS algorithm.

A new hybrid multibias analytical/decomposition-based FET parameter extraction algorithm with intelligent bias point selection

A new hybrid multibias analytical/decomposition-based parameter extraction procedure for GaAs FETs is described. The analytical calculations are integrated into an existing decomposition-based optimizer in a complementary approach, further increasing the robustness of the existing algorithm. It is illustrated that, in order to increase the reliability with which the full 15-element small-signal model can be extracted, it is necessary to exploit the underlaying characteristics of the system and the measured data used. This is achieved through the use of cold S-parameter data, along with simple modifications to the extraction algorithm, and a new intelligent selection algorithm for the active bias points used in the multi-bias extraction. The selection algorithm employs a simple geometric abstraction for the S-parameter data that allows it to select bias points that maximize the information available to the extraction procedure. The new selection algorithm shows for the first time what the influence of the bias points is on the performance of a multibias extraction procedure. Experimental results proving the robustness and accuracy of the described procedures are presented.

A comparison between hidden Markov models and vector quantization for speech independent speaker recognition

We compare Vector Quantization and Hidden Markov Models for speaker recognition for real time recognition. A scheme to reject speakers not known to the system is described and tested. Results show that the HMM algorithm outperforms the VQ algorithm. Using a 64 state HMM, a speaker recognition accuracy of 96.1% was achieved. The rejection option generated 25.7% false rejections for a 95% confidence of a correct decision. VQ best results were 93.1% with a 61% false rejection rate for codebooks of size 128.<<ETX>>

Algorithms for high order hidden Markov modelling

We detail an algorithm that transforms any higher order hidden Markov model (HMM) to an equivalent first order HMM. This makes it possible to process higher order HMMs with standard techniques applicable to first order models. Based on this equivalence, a fast incremental algorithm is developed for training higher order HMMs from lower order approximations, thereby avoiding the training of redundant parameters. This makes training of high order HMMs practical for many applications.We detail an algorithm that transforms any higher order hidden Markov model (HMM) to an equivalent first order HMM. This makes it possible to process higher order HMMs with standard techniques applicable to first order models. Based on this equivalence, a fast incremental algorithm is developed for training higher order HMMs from lower order approximations, thereby avoiding the training of redundant parameters. This makes training of high order HMMs practical for many applications.

A pseudo-skeletonization algorithm for static handwritten scripts

This paper describes a skeletonization approach that has desirable characteristics for the analysis of static handwritten scripts. We concentrate on the situation where one is interested in recovering the parametric curve that produces the script. Using Delaunay tessellation techniques where static images are partitioned into sub-shapes, typical skeletonization artifacts are removed, and regions with a high density of line intersections are identified. An evaluation protocol, measuring the efficacy of our approach is described. Although this approach is particularly useful as a pre-processing step for algorithms that estimate the pen trajectories of static signatures, it can also be applied to other static handwriting recognition techniques.

Calculation of LPC-based cepstrum coefficients using mel-scale frequency warping

This paper describes a method for warping the frequency axis of cepstrum coefficients in a way analogous to the preprocessing performed by the human ear. The equations are derived and historical background relating to different warping scales is discussed. The calculation is a two-step procedure in which the bilinear transform is used to represent the LPC coefficients on a warped frequency scale. A warping constant determines the degree of transformation. This results in an ARMA representation of the filter transfer function. The second step determines recursively the cepstrum coefficients corresponding to this ARMA transfer function. >This paper describes a method for warping the frequency axis of cepstrum coefficients in a way analogous to the preprocessing performed by the human ear. The equations are derived and historical background relating to different warping scales is discussed. The calculation is a two-step procedure in which the bilinear transform is used to represent the LPC coefficients on a warped frequency scale. A warping constant determines the degree of transformation. This results in an ARMA representation of the filter transfer function. The second step determines recursively the cepstrum coefficients corresponding to this ARMA transfer function.<<ETX>>

Verification of dynamic curves extracted from static handwritten scripts

Static handwritten scripts originate as images on documents and do not, by definition, contain any dynamic information. To improve the accuracy of static handwriting recognition systems, many techniques aim to estimate dynamic information from the static scripts. Mostly, the pen trajectories of the scripts are estimated. However, the efficacy of the resulting pen trajectories are rarely evaluated quantitatively. This paper proposes a protocol for the objective evaluation of automatically determined pen trajectories. A hidden Markov model is derived from a ground-truth trajectory. An estimated trajectory is then matched to the derived model. Statistics describing substitution, insertion and deletion errors are then computed from this match. The proposed algorithm is especially useful for performance comparisons between different pen trajectory estimation algorithms.