Cepstral Analysis

This paper introduces the use of genetics-based algorithm in the reduction of 24 parameter set (i.e the base set) to a 5,6,7,8 or 10 parameter set, for each speaker in text-independent speaker identification. The feature selection is done by finding the best features that discriminates a person from his/her two closest neighbors. The experimental results show that there is approsimately 5% increase in the recognition rate when the reduced set of parameters are used. Also the amount of calculation necessary for speaker recognition using the reduced set of features is much less than the amount of calculation required using the complete feature set in the testing phase. Hence it is more desirable to use the subset of the complete feature set found using the genetic algorithm suggested.

In this work, accurate spectral envelope estimation is applied to Voice Conversion in order to achieve High-Quality timbre conversion. True-Envelope based estimators allow model order selection leading to an adaptation of the spectral features to the characteristics of the speaker. Optimal residual signals can also be computed following a local adaptation of the model order in terms of the F0. A new perceptual criteria is proposed to measure the impact of the spectral conversion error. The proposed envelope models show improved spectral conversion performance as well as increased converted-speech quality when compared to Linear Prediction.

Moving shadows constitute problems in various applications such as image segmentation and object tracking. The main cause of these problems is the misclassification of the shadow pixels as target pixels. Therefore, the use of an accurate and reliable shadow detection method is essential to realize intelligent video processing applications. In this paper, a cepstrum-based method for moving shadow detection is presented. The proposed method is tested on outdoor and indoor video sequences using well-known benchmark test sets. To show the improvements over previous approaches, quantitative metrics are introduced and comparisons based on these metrics are made.

The article reports on the upgrading of the FPGA based isolated word recognition system for real-time tasks. All recognition system components (except some feature calculation steps) were implemented using VHDL. Some high precision calculations were implemented on soft core processor. The employed Dynamic time warping algorithm was speeded-up 2.8 times by restricting the calculated error matrix size. This enabled us to reduce the average word recognition time to 12.81 ms. Linear predictive coding, linear predictive coding cepstral and linear frequency cepstral coefficients feature analyses were investigated for 100 Lithuanian word recognition. In speaker dependent experiments linear predictive coding cepstral analysis gave the highest average recognition rate of 95 % and the highest robustness to white noise in speech. 15 dB noise level lowered average recognition rate to 86.2 %. Index Terms-Cepstral analysis, dynamic time warping, field programmable gate array, intellectual property core, isolated word recognition, linear predictive coefficients. Despite of recent software-based Lithuanian speech recognition [1] and synthesis [2] implementations on personal computers and servers there is an unaddressed need of embedded systems for mobile and stand-alone devices, interactive voice controlled systems, disabled person equipment, etc. Embedded systems bring in their specific requirements for speech recognizers: the limited speed of processing and of memory, the low power consumption. The recognition of large vocabulary and continuous speech requires complicated algorithms with huge amounts of calculations, large quantities of memory [3], [4]. This can result in enlarged power consumption, longer recognition time and higher recognition error rate. Many automatic speech recognition systems for the languages of minor use are now developed. Presented in Croatian speech recognizer uses acoustic models based on context-dependent triphone hidden Markov models (HMM) and phonetic rules. Experimentally it is shown that the system can be used for speech recognition with word error rate below 5 %. In [6] a speaker independent speech Manuscript

This article proposes the design of an automatic classifier using the empirical mode decomposition (EMD) along with machine learning techniques for identifying the five most important types of events of the Ubinas volcano, the most active volcano in Peru. The proposed method uses attributes from temporal, spectral, and cepstral domains, extracted from the EMD of the signals, as well as a set of preprocessing and instrument correction techniques. Due to the fact that multichannel sensors are currently being installed in seismic networks worldwide, the proposed approach uses a multichannel sensor to perform the classification, contrary to the usual approach of the literature of using a single channel. The presented method is scalable to use data from multiple stations with one or more channels. The principal component analysis method is applied to reduce the dimensionality of the feature vector and the supervised classification is carried out by means of several machine learning algorithms, the support vector machine providing the best results. The presented investigation was tested with a large database that has a considerable number of explosion events, measured at the Ubinas volcano, located in Arequipa, Peru. The proposed classification system achieved a success rate of more than 90%.

Emotion detection is a new research era in health informatics and forensic technology. Besides having some challenges, voice based emotion recognition is getting popular, as the situation where the facial image is not available, the voice is the only way to detect the emotional or psychiatric condition of a person. However, the voice signal is so dynamic even in a short-time frame so that, a voice of the same person can differ within a very subtle period of time. Therefore, in this research basically two key criterion have been considered; firstly, this is clear that there is a necessity to partition the training data according to the emotional stage of each individual speaker. Secondly, rather than using the entire voice signal, short time significant frames can be used, which would be enough to identify the emotional condition of the speaker. In this research, Cepstral Coefficient (CC) has been used as voice feature and a fixed valued kmeans clustered method has been used for feature classification. The value of k will depend on the number of emotional situations in human physiology is being an evaluation. Consequently, the value of k does not necessarily consider the volume of experimental dataset. In this experiment, three emotional conditions: happy, angry and sad have been detected from eight female and seven male voice signals. This methodology has increased the emotion detection accuracy rate significantly comparing to some recent works and also reduced the CPU time of cluster formation and matching.

Usually the mel-frequency cepstral coefficients are estimated either from a periodogram or from a windowed periodogram. We state a general estimator which also includes multitaper estimators. We propose approximations of the variance and bias of the estimate of each coefficient. By using Monte Carlo computations, we demonstrate that the approximations are accurate. Using the proposed formulas, the peak matched multitaper estimator is shown to have low mean square error (squared bias + variance) on speech-like processes. It is also shown to perform slightly better in the NIST 2006 speaker verification task as compared to the Hamming window conventionally used in this context.

This paper presents an overview of a state-of-the-art text-independent speaker verification system. First, an introduction proposes a modular scheme of the training and test phases of a speaker verification system. Then, the most commonly speech parameterization used in speaker verification, namely, cepstral analysis, is detailed. Gaussian mixture modeling, which is the speaker modeling technique used in most systems, is then explained. A few speaker modeling alternatives, namely, neural networks and support vector machines, are mentioned. Normalization of scores is then explained, as this is a very important step to deal with real-world data. The evaluation of a speaker verification system is then detailed, and the detection error trade-off (DET) curve is explained. Several extensions of speaker verification are then enumerated, including speaker tracking and segmentation by speakers. Then, some applications of speaker verification are proposed, including on-site applications, remote applications, applications relative to structuring audio information, and games. Issues concerning the forensic area are then recalled, as we believe it is very important to inform people about the actual performance and limitations of speaker verification systems. This paper concludes by giving a few research trends in speaker verification for the next couple of years.

W artykule przedstawiono sposób wyznaczania wskaźnika ekspozycji na niestacjonarne pola magnetyczne na podstawie adaptacyjnej analizy czasowo-częstotliwościowej, zarejestrowanych przebiegów czasowych indukcji pola magnetycznego B. Metodę adaptacyjną analizy czasowo-częstotliwościowej, opartą na algorytmie pogoni za dopasowaniem (ang. Matching Pursuit MP) zaimplementowano w wirtualnym analizatorze czas-częstotliwość, zaprojektowanym w graficznym środowisku programowania LabVIEW. Algorytm MP opiera się na redundantnym słowniku funkcji analizujących, z którego iteracyjnie wybierane są funkcje-atomy, najlepiej pasujące do składowych sygnału. Dzięki zmiennej długości okna oraz zmiennej częstotliwości modulacji algorytm MP pozwala na adaptacyjną, tzn. dopasowującą się do lokalnych struktur, reprezentację sygnału. Metoda adaptacyjna zastosowana do sygnałów pomiarowych, które są sumą ograniczonych w czasie przebiegów, występujących w różnych chwilach czasowych i mających różne pasma częstotliwości, ma zdecydowanie najlepszą łączną rozdzielczość czasowo-częstotliwościową w porównaniu z innymi kwadratowymi czasowo-częstotliwościowymi reprezentacjami sygnału. Spektrogram adaptacyjny nie zawiera również składowych interferencyjnych. Zaprezentowano przykładowe wyniki analiz małoczęstotliwościowego pola magnetycznego, zarejestrowanego na statku morskim. Słowa kluczowe: pomiary pola magnetycznego, analiza czasowo-częstotliwościowa, algorytm pogoni za dopasowaniem.

This article proposes the design of an automatic classifier using the empirical mode decomposition (EMD) along with machine learning techniques for identifying the five most important types of events of the Ubinas volcano, the most active volcano in Peru. The proposed method uses attributes from temporal, spectral, and cepstral domains, extracted from the EMD of the signals, as well as a set of preprocessing and instrument correction techniques. Due to the fact that multichannel sensors are currently being installed in seismic networks worldwide, the proposed approach uses a multichannel sensor to perform the classification, contrary to the usual approach of the literature of using a single channel. The presented method is scalable to use data from multiple stations with one or more channels. The principal component analysis method is applied to reduce the dimensionality of the feature vector and the supervised classification is carried out by means of several machine learning algorithms, the support vector machine providing the best results. The presented investigation was tested with a large database that has a considerable number of explosion events, measured at the Ubinas volcano, located in Arequipa, Peru. The proposed classification system achieved a success rate of more than 90%.

A new cepstrum-based channel compensation technique is proposed for speaker veri cation. Under this approach, channel cepstra are derived from the direct measurements of the frequency responses of telephone handsets. Speci cally, they are determined by truncating the inverse Fourier transform of the log magnitude of the interpolated handsets' frequency responses. The proposed method is readily applicable to telephonebased speaker veri cation. Experimental evaluations based on the telephone YOHO corpora suggest that the proposed channel compensation method strikes a good balance between veri cation performance and computational e ciency.

Preprint of the article: <strong><em>"Standardization of noisy volcano-seismic waveforms as a key step towards station-independent, robust automatic recognition" </em></strong> publshed in Seismological Research Letters-2019 (https://doi.org/10.1785/0220180334). This work is a 'deliverable' of the <em>VULCAN.ears</em> EC-funded project. This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie Grant Agreement No.[749249] (VULCAN.ears).

There is a paucity of quantitative data on the vestibular folds (VsF). Renewed interest in the VsF relates to their contribution to dysphonia. This study provides quantitative data to characterize developmentalinvolutional VsF changes and suggests a comprehensive description of their mechanics. Methods: 400 histological sections of 20 human larynges, matched by decade (1st-9th) were studied via light microscopy. Images were digitally captured with bright field and stereo microscopes. Linear and area measurements were made along with descriptions of constituent morphology. Results: 1. Thickness of the quadrangular membrane (QM) (epithelium and superficial lamina propria) was uniform (~10% of the area) along VsF length, consistent across ages, becoming fibrotic in old age. 2. Connective tissue (CT) was greatest in the mid-portion of VsF (40-45% of the total area) followed by posterior and anterior points of attachment. 3. Density/area of the mixed muco-serous glands (~ 50%) was slightly less in mid 1/3. 4. The densest CT area (~40%) was deep to the QM, had fimbriations surrounding nests of glandular tissue. In old age, these tissues became fibrotic and 5. increases in CT area accrued by addition of areolar tissue at the expanse of involuting gland acini.

Based on the Back Propagation Algorithm, this paper portrait a method for speaker identification in multiple foreign languages. In order to identify speaker, the complete process goes through recording of the speech utterances of different speakers in multiple foreign languages, features extraction, data clustering and system training. In order to realize the purpose, a database has been prepared which contains one sentence in 8 different international languages i.e. English spoken by 19 distinct speakers, both male and female, in each language. With total size of 760 speech utterances, the average performance of the system is 95.657%. Application of developed system is mainly used in speaker authentication in telephony security oriented applications where the normal conversations are of short durations and the tendency of the spokesperson is to switch language from one to another