Jacek Izydorczyk

A New Algorithm for Extraction of Parameters of Jiles and Atherton Hysteresis Model

Computation algorithm of the parameters of Jiles and Atherton (J-A) model of magnetic hysteresis has been briefly described in the paper. The basis of the algorithm is in the rational approximation of an unhysteretic curve of the material. The main advantages of the algorithm are: small amount of input data and simple computations which do not need extensive simulations or solving large systems of nonlinear equations

Speaker recognition system with good generalization properties

This paper presents speaker recognition system possessing very good generalization properties. Relatively low equal error rate for speaker verification and high identification rate for identification are achieved for very short training and testing sequences. This behaviour is achieved for the kernel modification of a classic Ho-Kashyap linear classifier. Achieved results for the new approach are compared with results for the classic GMM and VQ techniques. Speech of a moderately good quality from the Polish speech corpus was used for development of recognition system.

Influence of Feature Dimensionality and Model Complexity on Speaker Verification Performance

This paper provides description of a text dependent speaker recognition system based on vector quantization approach. The scope of this paper is to check influence of feature dimensionality and the complexity of the speaker model on verification process. Provided results show that MFCC features yield the lowest possible verification errors among all tested parameters. Although dimensionality of feature vectors is important, there is no need to increase it above some level as the improvement in verification performance is relatively low and computational complexity increases. Far more important than dimensionality is complexity of the speaker model.

Speech Recognition Based on Open Source Speech Processing Software

Creating of speech recognition application requires advanced speech processing techniques realized by specialized speech processing software. It is very possible to improve the speech recognition research by using frameworks based on open source speech processing software. The article presents the possibility of using open source speech processing software to construct own speech recognition application.

Influence of Corpus Size on Speaker Verification

The scope of this paper is to check influence of the size of the speech corpus on the speaker recognition performance. Obtained results for TIMIT corpus are compared with results obtained for smaller database ROBOT. Additionally influence of feature dimensionality and size of the speaker model was tested. Achieved results show that the best results can be obtained for MFCC features. The lowest EER for larger TIMIT database are 4 times worse than the best result for ROBOT corpus which confirms that biometric systems should be tested on as large data sets as possible to assure that achieved error rates are statistically significant.

Speaker Verification Performance Evaluation Based on Open Source Speech Processing Software and TIMIT Speech Corpus

Creating of speaker recognition application requires advanced speech processing techniques realized by specialized speech processing software. It is very possible to improve the speaker recognition research by using speech processing platform based on open source software. The article presents the example of using open source speech processing software to perform speaker verification experiments designed to test various speaker recognition models based on different scenarios. Speaker verification efficiency was evaluated for each scenario using TIMIT speech corpus distributed by Linguistic Data Consortium. The experiment results allowed to compare and select the best scenario to build speaker model for speaker verification application.

Invariant sums of higher order sensitivities

In this paper formulas for invariant sums of higher order sensitivities of voltage and current in the linear network are given. Considered network should be described by hybrid matrix. Formulas for invariant sums of sensitivities of first, second, third, and fourth order are presented.

Iterative Identification of Sparse Mobile Channels for TDS-OFDM Systems

Spectral efficiency improvement is essential for mobile systems development in the light of growing demand for high data transmission rate. The Time-domain synchronous OFDM transmission system increases bandwidth utilization by time-domain channel estimation without any support from frequency-domain pilot signals. Channel identification methods use a training sequence, added at the beginning of every OFDM symbol. Sparse nature of the channel impulse response, which manifests itself during wireless broadband transmission, enables one to implement compressed sensing techniques for channel identification. Channel sparsity reduces the number of measurement samples required in channel reconstruction procedure, but they should be taken from the inter-block interference (IBI) free part of the received training sequence. If maximum delay spread of the channel is comparable with the length of the training sequence, the number of measurements within the IBI-free region may be insufficient for accurate channel identification. This paper discusses a new iterative approach to sparse channels estimation, which uses the measurements from outside the IBI-free region. The proposed solution is examined for simulated transmission over randomly generated doubly-selective channels when compared with the conventional CS technique and perfect channel state information condition. The experiments reveal an enhancement of the transmission quality in terms of SER for slow-time selective channels when both delays and gains of the channel impulse response are estimated with the orthogonal matching pursuit algorithm. Further improvement, also for transmission over moderate varying channels, is observed when the path delays are found in advance. A suitable frame structure is proposed to do so.

Stopping Criteria Analysis of the OMP Algorithm for Sparse Channels Estimation

Wireless propagation environment utilised by broadband transmission systems usually has a sparse nature, i.e. only several isolated propagation paths are essential for information transfer. Receiver can recover the parameters of the particular paths using greedy, iterative algorithms that belong to the family of compressed sensing techniques. How to stop the iterative procedure, if no precise knowledge about the order of the channel sparsity is available in the receiver a priori, is a key question regarding a practical implementation of the method. The paper provides stopping criteria analysis of the Orthogonal Matching Pursuit (OMP) algorithm that is used as the core of the channel impulse response estimation method for Time-Domain Synchronous OFDM transmission system. There are investigated the residual error and the difference of successive residual errors of the OMP algorithm, as the possible metrics applied to stop the iteration procedure. Finally, a new heuristic stopping rule based on these two errors is proposed and numerically examined.

Three Steps to the Thermal Noise Death of Moore's Law

The purpose of this paper is to estimate the limits imposed on the computations made by logic gates fabricated in CMOS technology. It is assumed that the most rigorous constraints on computations originate from the thermodynamically irreversible processes that the computations consist of. In this paper, a model of CMOS logic gate is given based on the additive white Gaussian noise (AWGN) channel. It is found that the ultimate limit on energy spent to process one bit of data estimated by the proposed model of a CMOS logic gate is equal to the famous Landauer limit. This limiting energy cannot be obtained without sophisticated coding of the processed data. If one wants to avoid the processing of coded data, then much more energy has to be used to process one bit. In this paper, three limiting energies were estimated based on the proposed model of a logic gate. The limiting energies found in this study can be used, with Moores law, to predict time after which development of CMOS technology is not possible.