Search
displaying 1000 best results Help
Search results for: RECONSTRUCTION OF SPEECH SIGNALS
-
WYKORZYSTANIE SIECI NEURONOWYCH DO SYNTEZY MOWY WYRAŻAJĄCEJ EMOCJE
PublicationW niniejszym artykule przedstawiono analizę rozwiązań do rozpoznawania emocji opartych na mowie i możliwości ich wykorzystania w syntezie mowy z emocjami, wykorzystując do tego celu sieci neuronowe. Przedstawiono aktualne rozwiązania dotyczące rozpoznawania emocji w mowie i metod syntezy mowy za pomocą sieci neuronowych. Obecnie obserwuje się znaczny wzrost zainteresowania i wykorzystania uczenia głębokiego w aplikacjach związanych...
-
Investigating Feature Spaces for Isolated Word Recognition
PublicationMuch attention is given by researchers to the speech processing task in automatic speech recognition (ASR) over the past decades. The study addresses the issue related to the investigation of the appropriateness of a two-dimensional representation of speech feature spaces for speech recognition tasks based on deep learning techniques. The approach combines Convolutional Neural Networks (CNNs) and timefrequency signal representation...
-
Krzysztof Kutt dr inż.
PeopleComputer scientist and psychologist trying to combine expertise from both disciplines into something cool. My research activity focuses on the development of affective HCI/BCI interfaces (based on multimodal fusion of signals and contextual data), methods for processing sensory data (including semantization of such data) and the development of knowledge-based systems (in particular knowledge graphs and semantic web systems).
-
Detection of Lexical Stress Errors in Non-Native (L2) English with Data Augmentation and Attention
PublicationThis paper describes two novel complementary techniques that improve the detection of lexical stress errors in non-native (L2) English speech: attention-based feature extraction and data augmentation based on Neural Text-To-Speech (TTS). In a classical approach, audio features are usually extracted from fixed regions of speech such as the syllable nucleus. We propose an attention-based deep learning model that automatically de...
-
Zastosowanie spowalniania wypowiedzi w celu poprawy rozumienia mowy przez dzieci w szkole
PublicationThis paper presents a time-scale modification algorithms that could be used for hearing impairment therapy supported by real-time speech stretching. In this paper the OLA based algorithms and Phase Vocoder were described. In the experimental part usability of those algorithms for real-time speech stretching was discussed
-
KORPUS MOWY ANGIELSKIEJ DO CELÓW MULTIMODALNEGO AUTOMATYCZNEGO ROZPOZNAWANIA MOWY
PublicationW referacie zaprezentowano audiowizualny korpus mowy zawierający 31 godzin nagrań mowy w języku angielskim. Korpus dedykowany jest do celów automatycznego audiowizualnego rozpoznawania mowy. Korpus zawiera nagrania wideo pochodzące z szybkoklatkowej kamery stereowizyjnej oraz dźwięk zarejestrowany przez matrycę mikrofonową i mikrofon komputera przenośnego. Dzięki uwzględnieniu nagrań zarejestrowanych w warunkach szumowych korpus...
-
Instantaneous complex frequency for pipeline pitch estimation
PublicationIn the paper a pipeline algorithm for estimating the pitch of speech signal is proposed. The algorithm uses instantaneous complex frequencies estimated for four waveforms obtained by filtering the original speech signal through four bandpass complex Hilbert filters. The imaginary parts of ICFs from each channel give four candidates for pitch estimates. The decision regarding the final estimate is made based on the real parts of...
-
Creating new voices using normalizing flows
PublicationCreating realistic and natural-sounding synthetic speech remains a big challenge for voice identities unseen during training. As there is growing interest in synthesizing voices of new speakers, here we investigate the ability of normalizing flows in text-to-speech (TTS) and voice conversion (VC) modes to extrapolate from speakers observed during training to create unseen speaker identities. Firstly, we create an approach for TTS...
-
Janusz Smulko prof. dr hab. inż.
PeopleHe was born on April 25, 1964 in Kolno. He graduated in 1989 with honors from the Faculty of Electronics at Gdańsk University of Technology, specialising in measuring instruments. In 1989 he took second place in the Red Rose competition for the best student in the Pomerania Region. Since the beginning of his career ha has been associated with Gdańsk University of Technology: research assistant (1989-1996), Assistant Professor (1996-2012),...
-
PHONEME DISTORTION IN PUBLIC ADDRESS SYSTEMS
PublicationThe quality of voice messages in speech reinforcement and public address systems is often poor. The sound engineering projects of such systems take care of sound intensity and possible reverberation phenomena in public space without, however, considering the influence of acoustic interference related to the number and distribution of loudspeakers. This paper presents the results of measurements and numerical simulations of the...
-
Human voice modification using instantaneous complex frequency
PublicationThe paper presents the possibilities of changing human voice by modifying instantaneous complex frequency (ICF) of the speech signal. The proposed method provides a flexible way of altering voice without the necessity of finding fundamental frequency and formants' positions or detecting voiced and unvoiced fragments of speech. The algorithm is simple and fast. Apart from ICF it uses signal factorization into two factors: one fully...
-
PI observer stability and application in an induction motor control system
PublicationThe paper discusses the problem of stability of a proportional-integral Luenberger observer, designated for the state variables reconstruction of a linear, time-invariant dynamical system.
-
Investigating Feature Spaces for Isolated Word Recognition
PublicationThe study addresses the issues related to the appropriateness of a two-dimensional representation of speech signal for speech recognition tasks based on deep learning techniques. The approach combines Convolutional Neural Networks (CNNs) and time-frequency signal representation converted to the investigated feature spaces. In particular, waveforms and fractal dimension features of the signal were chosen for the time domain, and...
-
Strategie treningu neuronowego estymatora częstotliwości tonu krtaniowego z użyciem generatora syntetycznych samogłosek
PublicationW wielu zastosowaniach telekomunikacyjnych pojawia się problem przetwarzania lub analizy sygnału mowy, w ramach którego, często w obszarze podstawowych algorytmów, stosuje się estymator częstotliwości tonu krtaniowego. Estymator rozpatrywany w tej pracy bazuje na neuronowym klasyfikatorze podejmującym decyzje na podstawie częstotliwości oraz mocy chwilowej wyznaczanych w podpasmach analizowanego sygnału mowy. W pracy rozważamy...
-
Auditory-visual attention stimulator
PublicationNew approach to lateralization irregularities formation was proposed. The emphasis is put on the relationship between visual and auditory attention stimulation. In this approach hearing is stimulated using time scale modified speech and sight is stimulated by rendering the text of the currently heard speech. Moreover, displayed text is modified using several techniques i.e. zooming, highlighting etc. In the experimental part of...
-
INVESTIGATION OF THE LOMBARD EFFECT BASED ON A MACHINE LEARNING APPROACH
PublicationThe Lombard effect is an involuntary increase in the speaker’s pitch, intensity, and duration in the presence of noise. It makes it possible to communicate in noisy environments more effectively. This study aims to investigate an efficient method for detecting the Lombard effect in uttered speech. The influence of interfering noise, room type, and the gender of the person on the detection process is examined. First, acoustic parameters...
-
Audio-visual aspect of the Lombard effect and comparison with recordings depicting emotional states.
PublicationIn this paper an analysis of audio-visual recordings of the Lombard effect is shown. First, audio signal is analyzed indicating the presence of this phenomenon in the recorded sessions. The principal aim, however, was to discuss problems related to extracting differences caused by the Lombard effect, present in the video , i.e. visible as tension and work of facial muscles aligned to an increase in the intensity of the articulated...
-
Comparison of three methods of EPR retrospective dosimetry in watch glass
PublicationIn this article we present results of our follow-up studies of samples of watch glass obtained and examined within a framework of international intercomparison dosimetry project RENEB ILC 2021. We present three methods of dose reconstruction based on EPR measurements of these samples: calibration method (CM), added dose method (ADM) and added dose&heating method (ADHM). The study showed that the three methods of dose reconstruction...
-
Auditory Brainstem Responses recorded employing Audio ABR device
Open Research DataThe dataset consists of ABR measurements employing click, burst and speech stimuli. Parameters of the particular stimuli were as follows:
-
Surgical Site Infection after Breast Surgery: A Retrospective Analysis of 5-Year Postoperative Data from a Single Center in Poland
PublicationBackground and Objectives: Surgical site infection (SSI) is a significant complication of non-reconstructive and reconstructive breast surgery. This study aimed to assess SSI after breast surgery over five years in a single center in Poland. The microorganisms responsible for SSI and their antibiotic susceptibilities were determined. Materials and Methods: Data from 2129 patients acquired over five years postoperatively by the...
-
Ryzyko realizacji prac projektowych i prowadzenia przebudowy obiektów prywatnych
PublicationW artykule przedstawiono przykład realizacji przebudowy prywatnego budynku mieszkalnego, podczas której okazało się, że nie jest on wykonany zgodnie z projektem i zasadami sztuki budowlanej. Opracowany i zatwierdzony projekt budowlany przebudowy niewiele miał wspólnego z obiektem, którego dotyczył. Realizacja prac zgodnie z tym projektem była więc obarczona dużym ryzykiem, a w konsekwencji groziła katastrofą budowlaną.
-
Variable Ratio Sample Rate Conversion Based on Fractional Delay Filter
PublicationIn this paper a sample rate conversion algorithm which allows for continuously changing resampling ratio has been presented. The proposed implementation is based on a variable fractional delay filter which is implemented by means of a Farrow structure. Coefficients of this structure are computed on the basis of fractional delay filters which are designed using the offset window method. The proposed approach allows us to freely...
-
Method of reconstructing two-dimensional velocity fields on the basis of temperature field values measured with a thermal imaging camera
PublicationThis paper describes a novel numerical reconstruction procedure (NRP) of the velocity field during natural convective heat transfer from a two-sided, isothermal, heated vertical plate based only on the known temperature field obtained, e.g. with a thermal imaging camera. It has been demonstrated that with a knowledge of temperature distributions, the NRP enables the reconstruction of velocity fields by solving the Navier-Stokes...
-
Prof. Haitham Abu-Rub - A Visit to Poland's Gdansk University of Technology
PublicationReport on visit of Prof. Haitham Abu-Rub in Gdansk University of Technology. Speech on the Smart Grid Centre. Visit in the new smart grid laboratory of the GUT, the Laboratory for Innovative Power Technologies and Integration of Renewable Energy Sources (LINTE^2).
-
Mispronunciation Detection in Non-Native (L2) English with Uncertainty Modeling
PublicationA common approach to the automatic detection of mispronunciation in language learning is to recognize the phonemes produced by a student and compare it to the expected pronunciation of a native speaker. This approach makes two simplifying assumptions: a) phonemes can be recognized from speech with high accuracy, b) there is a single correct way for a sentence to be pronounced. These assumptions do not always hold, which can result...
-
Wikipedia Articles Representation with Matrix'u
PublicationIn the article we evaluate different text representation methods used for a task of Wikipedia articles categorization. We present the Matrix’u application used for creating computational datasets ofWikipedia articles. The representations have been evaluated with SVM classifiers used for reconstruction human made categories.
-
Digital microcontroller for sonar waveform generator
PublicationGenerating sounding signals is essential for the operation of active sonar. The system should be highly reliable. This can be achieved through architecture, communication between the devices, and a well-designed and self–testing software. The system presented in the article is responsible for the generation of hydroacoustic sounding signals, and ensures proper interaction between power amplifiers and power supplies. Thanks to its...
-
Surface EMG-based signal acquisition for decoding hand movements
Open Research DataBiosignal processing plays a crucial role in modern hand prosthetics. The challenge is to restore functionality of a lost limb based on the signals acquired from the surface of the stump. The number of sensors (emg channels) used for signal acquisition influence the quality of a prosthetic hand. Modern algorithms (including neural networks) can significantly...
-
Rediscovering Automatic Detection of Stuttering and Its Subclasses through Machine Learning—The Impact of Changing Deep Model Architecture and Amount of Data in the Training Set
PublicationThis work deals with automatically detecting stuttering and its subclasses. An effective classification of stuttering along with its subclasses could find wide application in determining the severity of stuttering by speech therapists, preliminary patient diagnosis, and enabling communication with the previously mentioned voice assistants. The first part of this work provides an overview of examples of classical and deep learning...
-
A comparative study of English viseme recognition methods and algorithm
PublicationAn elementary visual unit – the viseme is concerned in the paper in the context of preparing the feature vector as a main visual input component of Audio-Visual Speech Recognition systems. The aim of the presented research is a review of various approaches to the problem, the implementation of algorithms proposed in the literature and a comparative research on their effectiveness. In the course of the study an optimal feature vector...
-
Modeling and Designing Acoustical Conditions of the Interior – Case Study
PublicationThe primary aim of this research study was to model acoustic conditions of the Courtyard of the Gdańsk University of Technology Main Building, and then to design a sound reinforcement system for this interior. First, results of measurements of the parameters of the acoustic field are presented. Then, the comparison between measured and predicted values using the ODEON program is shown. Collected data indicate a long reverberation...
-
A comparative study of English viseme recognition methods and algorithms
PublicationAn elementary visual unit – the viseme is concerned in the paper in the context of preparing the feature vector as a main visual input component of Audio-Visual Speech Recognition systems. The aim of the presented research is a review of various approaches to the problem, the implementation of algorithms proposed in the literature and a comparative research on their effectiveness. In the course of the study an optimal feature vector construction...
-
Comparative analysis of various transformation techniques for voiceless consonants modeling
PublicationIn this paper, a comparison of various transformation techniques, namely Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT) and Discrete Walsh Hadamard Transform (DWHT) are performed in the context of their application to voiceless consonant modeling. Speech features based on these transformation techniques are extracted. These features are mean and derivative values of cepstrum coefficients, derived from each transformation....
-
Multistatyczny, Dopplerowski System określania położenia i prędkości ruchomych celów w wodzie
PublicationW omawianym w pracy multistatycznym, dopplerowskim systemie określania położenia i prędkości ruchomych celów w wodzie źródłem sygnału są dwa nadajniki emitujące sinusoidalne, akustyczne fale ciągłe o różnych częstotliwościach, które po odbiciu od ruchomego celu są obierane przez cztery hydrofony. W artykule przedstawiono analize teoretyczna efektu Dopplera, na którym oparte jest działanie systemu oraz metodę rozwiązania głównych...
-
Applicability of null-steering for spoofing mitigation in civilian GPS
PublicationCivilian GPS signals are currently used in many critical applications, such as precise timing for power grids and telecommunication networks. Spoofing may cause their improper functioning. It is a threat which emerges with the growing availability of GPS constellation simulators and other devices which may be used to perform such attack. Development of the effective countermeasures, covering detection and mitigation, is necessary...
-
Szybka identyfikacja harmonicznych na podstawie oszczędnego próbkowania
PublicationW pracy przedstawiono implementację szybkiego algorytmu rekonstrukcji sygnału, opartego na teorii oszczędnego próbkowania, który może wykrywać harmoniczne w sygnale wejściowym. Zagadnienie rekonstrukcji sygnału jest problemem optymalizacyjnym rozwiązywanym za pomocą algorytmu programowania liniowego. Dodatkowo, aby przyspieszyć zbieżność rozwiązania zastosowano w rzadkiej dziedzinie sygnału filtr typu K-rank-order. Przeprowadzona...
-
Comparison of near infrared spectroscopy (NIRS) and near-infrared transillumination-backscattering sounding (NIR-T/BSS) methods
PublicationThe aim of the study was to compare simultaneously recorded a NIR-T/BSS and NIRS signals from healthy volunteers. NIR-T/BSS is a device which give an ability to non-invasively detect and monitor changes in the subarachnoid space width (SAS). Experiments were performed on a group of 30 healthy volunteers (28 males and 2 females, age 30.8 ± 13.4 years, BMI = 24.5 ± 2.3 kg/m2). We analysed recorded signals using analysis methods based...
-
A Wearable System Developed to Monitor People Suffering from Vasovagal Syncope
PublicationA wearable system for monitoring non-invasively signals invaluable when examining person suffering from vasovagal syncope is presented in the paper. Following signals are continuously recorded: electrocardiogram, photopletysmogram, impedance cardiogram and electrodermal resistance.
-
Rzadka reprezentacja sygnału niestacjonarnego w technice oszczędnego próbkowania
PublicationPrzedstawiono zastosowanie techniki oszczędnego próbkowania do rekonstrukcji sygnału niestacjonarnego na podstawie skompresowanych próbek w dziedzinie czas-częstotliwość. Zastosowano nadmiarowy algorytm z różnymi słownikami aby znaleźć rzadką reprezentację sygnału. Wyniki symulacji potwierdzają, że zastosowanie oszczędnego próbkowania pozwala na rekonstrukcję sygnału niestacjonarnego z małej liczby losowo pobranych próbek, z niewielką...
-
Reliability of Pulse Measurements in Videoplethysmography
PublicationReliable, remote pulse rate measurement is potentially very important for medical diagnostics and screening. In this paper the Videoplethysmography was analyzed especially to verify the possible use of signals obtained for the YUV color model in order to estimate the pulse rate, to examine what is the best pulse estimation method for short video sequences and finally, to analyze how potential PPG-signals can be distinguished from...
-
Self diagnostics using smart glasses - preliminary study
Publicationn this preliminary study we analyzed the possibility of the reliable measurement of biomedical signals with some potential hardware extensions of smart glasses. Using specially designed experimental prototypes four category of biomedical signals were measured: electrocardiograms, electromyograms, electroencephalograms and respiration waveforms. Experi- ments with volunteers proved that using even simple construc- tion of sensors...
-
Receiver of Doppler multistatic system for moving target detection and tracking
PublicationThe article presents a method for solving major structural problems that occur in the receiver used in the multistatic Doppler system, aimed at determination of the trajectory and velocity of a moving target. In the system two transmitters emit acoustic continuous sinusoidal waves at different frequencies. The signals, scattered from a moving target are received by four hydrophones. Beside of the echoes, much larger signals coming...
-
The dynamic signature verification using population-based vertical partitioning
PublicationThe dynamic signature is an attribute used in behavioral biometrics for verifying the identity of an individual. This attribute, apart from the shape of the signature, also contains information about the dynamics of the signing process described by the signals which tend to change over time. It is possible to process those signals in order to obtain descriptors of the signature characteristic of an individual user. One of the methods...
-
A multisensor detector of a sleep apnea for using at home
PublicationDiagnosis of obstructive sleep apnea usually involves polysomnographic analysis, which unfortunately requires overnight stay in a specialized clinic and is very uncomfortable for a patient. This paper describes the method and apparatus for recording a set of signals to detect sleep apnea. The device records the following signals simultaneously: three-channel ECG, respiratory functions, signals from the accelerometer, and snoring...
-
Effectiveness of the robust PSS design
PublicationThe paper discusses optimal PSS of synchronous generator synthesis. The optimal controller is an Hinf controller, what means that minimises Hinf norm of transfer function between the exogenous signals such as reference inputs and disturbances, and the error signals which are to be minimised to meet the control objective. The dynamic properties of the plant are shaped by choosing appropriate weighting function applied to the plant...
-
SkinDepth - synthetic 3D skin lesion database
Open Research DataSkinDepth is the first synthetic 3D skin lesion database. The release of SkinDepth dataset intends to contribute to the development of algorithms for:
-
Aleksander Mroziński dr inż.
PeopleAleksander Mroziński is a PhD student at the Faculty of Electronics, Telecommunications and Informatics of the Gdańsk University of Technology. From June 2019 is employed as a metadata editor in the project THE BRIDGE OF DATA. He also participates in three other projects implemented at the Gdańsk University of Technology: InterPhD II, OPUS 13 and PROM. His desire to discover the world motivates him to acquire multidisciplinary...
-
Zastosowanie sygnałów o projektowanych kształtach do diagnostyki obiektów wysoko-impedancyjnych metodą spektroskopii impedancyjnej
PublicationW artykule przedstawiono metodę szybkiej spektroskopii impedancyjnej obiektów o wysokich impedancjach (|Zx| > 1 GOhm) z zastosowaniem sygnałów o projektowanych kształtach. Sygnał pobudzenia wytwarzany jest w module DAQ U2531A i doprowadzany na wejście badanego obiektu za pośrednictwem przetwornika cyfrowo-analogowego (CA). Sygnały odpowiedzi proporcjonalne do napięcia na mierzonej impedancji Zx oraz prądu płynącego przez Zx są...
-
Analiza techniczno-ekonomiczna rozwiązań rewitalizacyjnych zabytkowego mostu w Tczewie
PublicationPrzykłady koncepcji odbudowy przyczółka zachodniego zabytkowego mostu przez rzekę Wisłę w Tczewie wraz z zespołem bramnym. Ocena propozycji trzech rozwiązań rewitalizacyjnych mostu i wybór optymalnego pod względem finansowym i technologicznym. Czas i koszty budowy oraz wierność odbudowy zespołu bramnego mostu jako kryteria optymalizacji. Analiza w celu pozyskania środków finansowych potrzebnych na odbudowę, ale także adaptację...
-
The accuracy of pulse rate estimation from the sequence of face images
PublicationThe goal of this paper is to analyze the accuracy of pulse rate estimation from the sequence of face images. Simulated and real signals were used to evaluate two pulse rate estimators; one for frequency domain and the second one for time domain using the autocorrelation function. The results show that the mean difference between the reference measurements and estimated pulse rate values are about 2bpm. In the analysis of short...