Voice Signal Research Articles

Secure Communication of Fractional Complex Chaotic Systems Based on Fractional Difference Function Synchronization

Fractional complex chaotic systems have attracted great interest recently. However, most of scholars adopted integer real chaotic system and fractional real and integer complex chaotic systems to improve the security of communication. In this paper, the advantages of fractional complex chaotic synchronization (FCCS) in secure communication are firstly demonstrated. To begin with, we propose the definition of fractional difference function synchronization (FDFS) according to difference function synchronization (DFS) of integer complex chaotic systems. FDFS makes communication secure based on FCCS possible. Then we design corresponding controller and present a general communication scheme based on FDFS. Finally, we respectively accomplish simulations which transmit analog signal, digital signal, voice signal, and image signal. Especially for image signal, we give a novel image cryptosystem based on FDFS. The results demonstrate the superiority and good performances of FDFS in secure communication.

Health Parameter Analyses of Living Organism in Underwater Environment

Health monitoring of any living organism can be done easily through various wireless techniques in air medium. Whereas in the water medium, it is very difficult to monitor the health parameter of the living animals such as fish, rare species like Dolphins and even the human beings working underwater. This is because of the high attenuation and severe multipath effects. To overcome this in this paper, a dolphin signal is used as carrier signal for communication purpose. The error rate is calculated and compared with the standard modulation scheme Binary Phase Shift Keying (BPSK). The simulation results evident that the underwater communication by the low frequency signals like human voice and Dolphin signal performs better and gives an error rate equals to the one-tenth of the error rate caused by the BPSK scheme in water medium.

The struggle for the range of telephone communication before the invention of the audion Lee de Forest

The increase in the range of laying telephone lines at the beginning of the 20th century raised the issue of high-quality telephone signal transmission for scientists and engineers. The innovative technology for solving the problem was based on the “Heaviside condition”. AT&T engineer George Campbell, who received a patent for the design of load coils, was engaged in this task. He demonstrated that a telephone line loaded with coils can transmit clear voice signals twice as far as unloaded telephone lines. At about the same time, Columbia University professor Michael Pupin dealt with this problem, who also received a patent for a similar concept in priority over Campbell's application. As a result, load coils became known as “Pupin coils”. Ironically, Campbell lost the patent priority battle. The proposed innovation eliminated time delays and distortions in the signal, thereby significantly increasing the transmission speed.

Identification of People with Parkinson's Suspicions through Voice Signal Processing

Parkinson is considered a disease with a very random prognosis, in addition to its origin due to a multisystemic neurodegenerative process that affects the central nervous system, which is responsible for motor control of the body and also produces chronic joint pain if the patient is not treated also suffers states of depression. This disease currently has no cure so it recommends the patient's family to provide quality of life, the age of incidence is from 40 years, according to the INCN (Instituto Nacional de Ciencias Neurologicas) indicates that there are 3,000 cases of Parkinson in Peru annually. In this research paper, it proposes the creation of an algorithm in MATLAB capable of extracting the characteristics of the voice spectrum through the voice signal processing to provide an early detection so that they can receive treatment, appease and slow down Parkinson's disease. This processing will consist of submitting the audio by the Fast Fourier Transform (FFT), identifying the signal bodies, separating by frequency periods, to finally find the average and maximum values. It was identified that in the lower frequencies are where there are major differences, in addition the test was done with patients who has Parkinson's suspicions and the same differences were obtained resulting in the frequency periods [9Hz – 13Hz], [20Hz –30Hz] and [40Hz – 54Hz]. Also note that the period of 20 Hz to 30 Hz is where if the values in this frequency are less than 3.5 in amplitude they are principles of suspicion of Parkinson's disease.

Automated Detection of Parkinson's Disease Based on Multiple Types of Sustained Phonations Using Linear Discriminant Analysis and Genetically Optimized Neural Network.

Objective: Parkinson’s disease (PD) is a serious neurodegenerative disorder. It is reported that most of PD patients have voice impairments. But these voice impairments are not perceptible to common listeners. Therefore, different machine learning methods have been developed for automated PD detection. However, these methods either lack generalization and clinically significant classification performance or face the problem of subject overlap. Methods: To overcome the problems discussed above, we attempt to develop a hybrid intelligent system that can automatically perform acoustic analysis of voice signals in order to detect PD. The proposed intelligent system uses linear discriminant analysis (LDA) for dimensionality reduction and genetic algorithm (GA) for hyperparameters optimization of neural network (NN) which is used as a predictive model. Moreover, to avoid subject overlap, we use leave one subject out (LOSO) validation. Results: The proposed method namely LDA-NN-GA is evaluated in numerical experiments on multiple types of sustained phonations data in terms of accuracy, sensitivity, specificity, and Matthew correlation coefficient. It achieves classification accuracy of 95% on training database and 100% on testing database using all the extracted features. However, as the dataset is imbalanced in terms of gender, thus, to obtain unbiased results, we eliminated the gender dependent features and obtained accuracy of 80% for training database and 82.14% for testing database, which seems to be more unbiased results. Conclusion: Compared with the previous machine learning methods, the proposed LDA-NN-GA method shows better performance and lower complexity. Clinical Impact: The experimental results suggest that the proposed automated diagnostic system has the potential to classify PD patients from healthy subjects. Additionally, in future the proposed method can also be exploited for prodromal and differential diagnosis, which are considered challenging tasks.

Automatic Voice Pathology Monitoring Using Parallel Deep Models for Smart Healthcare

Recent advancements in wireless communication and machine learning technologies aid in the development of an accurate and affordable healthcare facility. In this paper, we propose a smart healthcare framework in a mobile platform using deep learning. In the framework, a smartphone records a voice signal of a client and sends it to a cloud server. The cloud server processes the signal and classifies it as normal or pathological using a parallel convolutional neural network model. The decision on the signal is then transferred to the doctor for a prescription. Two publicly available databases were used in the experiments, where voice samples were played in front of a smartphone. The experimental results show the suitability of the proposed framework in the healthcare framework.

Acoustic Tracking of Pitch, Modal, and Subharmonic Vibrations of Vocal Folds in Parkinson’s Disease and Parkinsonism

The prominent and early presence of dysphonia is considered a valuable marker for differentiation of idiopathic Parkinson’s disease and parkinsonian syndromes. Objective quantification of vibrational regimes represented by modal and subharmonic vibrations may thus be vital for improving accuracy of diagnostic decision. The rationale for analyzing vibrational regimes is that abnormal subharmonic vibrations might be the key factor causing dysphonia in parkinsonian syndromes. This study introduces a new fully automated methodology based on robust pitch tracker for decoupling vibrations controlled by laryngeal muscles from the effect of subharmonics that provides distinguishing features of Parkinson’s disease and atypical parkinsonian syndromes. We tested the method on resynthesized signals with known parameters and demonstrated that vibrations controlled by laryngeal muscles as well as subharmonics can be detected reliably with a precision that outperforms available technologies. We analysed 337 sustained vowels of 22 patients with PD, 21 patients with multiple system atrophy, 18 patients with progressive supranuclear palsy and 22 healthy controls. Our results showed that subharmonics are more prominent in atypical parkinsonian syndromes compared to Parkinson’s disease. Also, increased modulation by laryngeal muscles appears to be a distinctive symptom of multiple system atrophy. Developed algorithm and proposed resynthesized voice signals provide further critical step to understanding and evaluation of dysphonia in Parkinsonism.

Preliminary study of Data Mining analysis of high speed Kymography and voice data

The use of high-speed laryngeal images technology contributes increasingly to promote greater objectivity in the characterization of laryngeal physiology, as well as, in the diagnosis and monitoring of laryngeal diseases. Vocal fold paralysis is still an unknown incidence, both in Brazil and in the rest of the World, its occurrence has become more frequent in medical and speech-pathologist clinics. Data mining is the process of analyzing hidden patterns of data according to different perspectives for categorization into useful information. In other words, Data Mining is able to identify patterns between information and grouping them according to some given criteria. Objective: To analyze high-speed kymographic images and voice signals with and without unilateral vocal fold paralysis using the data mining computerized system DAMICORE. Methodology: High-speed kymography and acoustic voice signals of subjects with and without unilateral vocal fold paralysis were analyzed by a computerized data mining tool, the DAMICORE system. Results: Data mining analysis for High-speed kymography of healthy individuals gather the groups by sex, while images from subjects with unilateral vocal fold paralysis were gathered by contrast of colors and by light incidence in the images. For voice signals, the technique considered the presence of external noise as a criterion for gathering groups. Conclusion: We concluded that DAMICORE it is a promising tool for kymographic images and voice signal data mining analysis. This tool shown high sensibility to noise in acoustical signals. Consequently, future research should consider this characteristic.

Parkinson Disease Detection Based on Voice and EMG Pattern Classification Method for Indonesian Case Study

Parkinson disease (PD) detection using pattern recognition method has been presented in literatures. This paper present multi-class PD detection utilizing voice and electromyography (EMG) features of Indonesian subjects. The multi-class classification consists of healthy control, possible stage, probable stage and definite stage. These stages are based on Hughes scale used in Indonesia for PD. Voice signals were recorded from 15 people with Parkinson (PWP) and 8 healthy control subjects. Voice and EMG data acquistion were conducted in dr Kariadi General Hospital Semarang, Central Java, Indonesia. Twenty two features are used for voice signal feature extraction and twelve features are emploed for EMG signal. Artificial Neural Network is used as classification method. The results of voice classification show that accuracy for testing step of 94.4%. For EMG classification, the accuracy of testing of 71%.

A Machine Learning Approach for Speech Detection in Modern Wireless Communication Environment

Modern wireless communication has gained a improved position as compared to previous time. Similarly, speech communication is the major focus area of research in respective applications. Many developments are done in this field. In this work, we have chosen the OFDM modulation based communication system, as it has importance in both licensed and unlicensed wireless communication platform. The voice signal is passed though the proposed model to obtain at the receiver end. Due to different circumstances, the signal may be corrupted partially at the user end. Authors try to achieve a better signal for reception using a neural network model of RBFN. The parameters are chosen for the RBFN model, as energy, ZCR, ACF, and fundamental frequency of the speech signal. In one part these parameters have eligibility to eliminate noise partially, where as in other part the RBFN model with these parameters proves its efficacy for both noisy speech signals with noisy channel as Gaussian channel. The efficiency of OFDM model is verified in terms of symbol error rate and the transmitted speech signal is evaluated in term of SNR that shows the reduction of noise. For visual inspection, a sample of signal, noisy signal and received signal is also shown. The experiment is performed with 5dB, 10dB, 15dB noise levels. The result proves the performance of RBFN model as the filter.The performance is measured as the listener’s voice in each condition. The results show that, at the time of the voice in noise environment, proposed technique improves the intelligibility on speech quality.

V-Speech

Smart glasses are often used in public environments or industrial scenarios that are relatively noisy. Background noise and sound from competing speakers deteriorate voice communication or performance of automatic speech recognition (ASR). Typically, signal processing techniques are used to reduce noise and enhance voice quality, but they have limitations in performance, hardware and/or computing resources. Voice capturing techniques using bone conducting on the head have been proposed in some experimental and commercial devices, with good robustness against environmental noise, but limited by signal distortions inherent to the capturing method. We present V-Speech, a novel sensing and signal processing solution that enables speech recognition and human-to-human communication in very noisy environments. It captures the voice signal with a vibration sensor located in the nasal pads of smart glasses and performs a transformation to the sensor signal in order to mimic that of a regular microphone in low noise conditions. The signal transformation is key, as it eliminates the "nasal distortion" that is introduced for nasal phonemes in the speech induced vibrations of the nasal bone. The output of V-Speech has low noise, sounds natural, and can be used in voice communication or as input to an off-the-shelf ASR service. We evaluated V-Speech in noise-free and noisy conditions with 30 volunteer speakers uttering 145 phrases and validated its performance on ASR engines and with assessments of voice quality using the Perceptual Evaluation of Speech Quality (PESQ) metric. The results show in extreme noise conditions a mean improvement of 50% for Word Error Rate (WER), and 1.0 on a scale of 5.0 for PESQ. In addition, real life recordings were made under various representative noise conditions, some with sound pressure levels of 93 dBA, which require hearing protection. Subjective listening tests were conducted according to a modified ITU P.835 approach to determine intelligibility, naturalness and overall quality. Under these extreme conditions, where V-Speech achieved 30 dB SNR, subjective results show the speech is intelligible, and the naturalness of the speech is rated as fair to good. This enables clear voice communication in challenging work environments, for example in places with industrial, factory, mining and construction noise. With our proposed smart switching technique between a regular microphone signal and V-Speech, the optimal quality can be maintained from low to high noise conditions.

The Impact of Nasalance on Cepstral Peak Prominence and Harmonics‐to‐Noise Ratio

Objectives/HypothesisCepstral peak prominence (CPP) has been reported as a reliable measure of dysphonia and a preferred alternative to harmonics‐to‐noise ratio (HNR). However, CPP has been observed to be sensitive to articulatory variation and vocal intensity. The aim of this study was to examine the impact of nasalance on CPP and HNR of voice signals. It was hypothesized that increased nasalance would be associated with decreased CPP.Study DesignWithin‐subject correlation design.MethodsThirty vocally healthy female participants were recorded reading and producing a vowel in alternation with a nasal consonant while wearing a nasometer for calculation of nasalance. Recorded vowel, nasalized, and nasal segments of speech were used to calculate CPP using Analysis of Dysphonia in Speech and Voice software, and HNR and vocal intensity using Praat software.ResultsSignificant main effects of conditions were observed for CPP. CPP values decreased significantly when phonation changed from vowel to nasalized vowel and to nasal. There was correlation between CPP and nasalance and between CPP and intensity. HNR was slightly higher in the nasal condition than in vowel. There was a weak correlation between HNR and nasalance. No correlation was found between HNR and intensity.ConclusionsCPP is sensitive to changes in vocal tract configuration caused by nasalization as well as intensity, whereas HNR is not. Therefore, CPP may reflect the periodicity in source signal or the filtering effects of vocal tract. Further research is needed to clarify the application and interpretation of CPP in clinical practice.Level of Evidence4 Laryngoscope, 129:E299–E304, 2019

Research of the informativeness the phase data of the user voice signal of the authentication system

Ways to improve the efficiency of modern voice authentication systems in various access systems are investigated in the article. One of the main ways to increase the effectiveness of authentication systems under consideration is associated with the use of voice signal phase. The object of the research is the process of digital signal processing in voice authentication systems. The scientific problem of forming and using phase data of a voice signal of an au-thentication system is solved. The purpose of the research is to evaluate the informative value of phase data of a voice signal and determine their main characteristics. The formant information on the amplitude spectrum of the experimental voice signal was processed and its qualitative and quantitative characteristics were obtained. Then, the phase data of the analyzed signal were calculated, the features of their use were revealed, and the phase spectrum was obtained on the basis of the calculation results. Processing of the phase spectrum has shown that it is easier to select at least one and a half times bigger formant of a voice signal that can be used to authenticate a user. The approximation of the maxima of formants is performed using linear and quadratic polynomials. The qualitative and quantitative evaluation of the amplitude and phase spectra formant information has confirmed the hypothe-sis that the phase data are more informative. The presented research results should be used in voice authentication systems, in improving speech recogni-tion systems, as well as in solving speaker identification problems.

Signal denoising optimization based on a Hilbert-Huang transform-triple adaptable wavelet packet transform algorithm

The Hilbert-Huang transform (HHT) can retain intrinsic signal characteristics after noise reduction but still leaves a slightly noisy signal, and the wavelet packet transform (WPT) denoising algorithm eliminates noise efficiently but causes distortion of the original signal. To overcome these issues, this paper proposes to combine these two algorithms linearly to maximize the signal-to-noise ratio (SNR) and increase the adaptive optimal solution for the three main steps involved in the WPT. The proposed algorithm is tested on voice signals with different background noise intensities and different noise functions in order to test the robustness of the new Hilbert-Huang transform triple adaptable wavelet packet transform (HHT-TAWPT) algorithm. The results prove that the proposed algorithm effectively denoises the signal while keeping the original signal intact and this was indicated by the segmental SNR and frequency spectrograms when compared to the individual HHT and WPT algorithms.

Voice Control of Multiple Devices Based on Speech

Abstract This paper presents a system that employs the use of speech technology to control electronic devices attached to PC. The work in this paper is implemented in two major phases. Phase one is the software part of the system. Its role is to accept voice signal from a microphone attached to the PC and perform speech recognition on the signal, determine operating commands from the recognized phrases and control devices attached to the computer ports. Phase two is hardware logic circuit connected between the PC’s printer port and the devices being controlled. This logic circuit which is implemented using simple logic gates performs addressing and control mode for the electronic devices being controlled. The system runs on windows vista operating system and it is capable of authenticating users and validating their action. User response for the system is provided in both voice response and graphic display messages.

The Design and Implementation of the Intelligent Warehouse Voice Control Guard System Based on Voiceprint Recognition

This design is an intelligent warehouse voice control system based on voiceprint recognition. The system consists of the following components: the first switch and the second switch connected via LAN; the voice control guard of the first circuit-connected switch, which collects voice signals on site and outputs voiceprint information; the database server of the second circuit-connected switch, which compares voiceprint information with the voiceprint samples in the database server for voiceprint recognition; the access control switch module, which receives access control commands and then opens or closes the door. Based on FPGA, infrared sensing technology, voice acquisition technology, human-computer interaction technology, and data transmission and communication technology, this design is capable of recognizing the identification of personnel and performing authentication management. This design also features convenient voice acquisition, low cost, compact size, various functions, and strong scalability.

Methodologies of speech analysis for neurodegenerative diseases evaluation

Background and objectiveNeurodegenerative diseases are disorders that affect neurons in the brain resulting in a debilitating condition and progressive degeneration of nerve cells. These diseases involve different aspects among which speech impairment. Vocal signal analysis is used to evaluate this impairment and to discriminate normal from pathological voices. Materials and methodsIn this paper, two methods of vocal signal analysis have been proposed to evaluate an anomalous condition in human speech, known as dysarthria, useful to compare pathological and healthy voices. Parkinson and Multiple Sclerosis disease have been considered and patients affected by both pathologies have been enrolled. The methods have been tested on 153 voice signals belonging to: 39 healthy subjects (HS), 60 patients with Parkinson's Disease (PD) and 54 patients with Multiple Sclerosis (MS). Acoustic (F0, jitter, shimmer, NHR) and vowel metric (tVSA, qVSA, FCR) features have been extracted. ResultsThe results report significant differences in almost all of these features in pathological and healthy voices by performing statistical tests. F0, jitter, shimmer, NHR, tVSA and FCR are statistically significant features thus they can be used as indicators in the diagnosis of dysarthria-related diseases such as in PD and MS. The results suggest that the applied methodologies are efficient and useful in characterizing the different behavior of vocal signal in healthy and pathological subjects. Consequently, they could be a valid support for physicians in disease evaluation and progression monitoring. ConclusionsThe contribution aims to evaluate, support and diagnose the comorbidity in pathological patients verifying the co-occurrence of speech and neurological disorders in the same individual. The proposed solution is studied and implemented to be efficient and low cost following the model of precision medicine to customize clinical practice in disease diagnosis and treatment.

A Two-Stage Cepstral Analysis Procedure for the Classification of Rough Voices

The objective of this study was to investigate the ability of a two-stage method of cepstral peak identification to effectively discriminate rough vs breathy vs typical voice in sustained vowel productions. It was hypothesized that a dual-stage search for cepstral peak prominences (CPP's) above and below specified quefrency/F0 cutoffs would result in a CPP difference that would be characteristic of the rough, diplophonic voice type. Central one-second portions of sustained vowel /a/ productions were obtained from 90 subjects (rough, breathy, and normophonic voices). All voice samples were analyzed using a a two-stage cepstral analysis process in which a CPPHigh-Low difference value was obtained by identifying cepstral peaks above and below a lower limit for expected F0 (150Hz for females and 90Hz for males), called CPPHigh and CPPLow respectively. The CPPHigh-Low difference value was observed to be a highly significant predictor, with negative values for this parameter characteristic of a dominant subharmonic in the voice signal and the perception of diplophonic, rough voice. Correct classification of rough vs nonrough voice samples was 82.2% (sensitivity 0.80 and specificity 0.833). In the consideration of three group classification (breathy vs. normophonic vs. rough), models incorporating two predictors (the CPP obtained from a single search through a 60 to 300Hz frequency range (CPPDefault) and the CPPHigh-Low difference value) correctly classified 78.88% of the voice samples. Rough, diplophonic voices were consistently observed to have a subharmonic peak that was greater in amplitude than the cepstral peak obtained within the region of the expected F0, resulting in a negative value for the CPPHigh-Low difference. The two-stage cepstral analysis process described herein is visually intuitive from the graphical display of a cepstrum and is a simple extended calculation derived from cepstral analysis procedures that have been recommended as essential in the acoustic description of vocal quality.

Sub-band coding for audio signals using Matlab

This work presents the results of a research project aimed at studying the coding of audio signals in order to obtain a compact and efficient digital representation for transmission or storage purposes. A psychoacoustic model based on sub-band coding is implemented in MATLAB®, which identifies the type of audio: voice or music signal, based on a model of the human auditory system. Our focus was on measuring the relationship between reduced transmission/storage bitrate and audio quality.

Arabic discourse analysis based on acoustic, prosodic and phonetic modeling: elocution evaluation, speech classification and pathological speech correction

This work describes a complete study addressing the pathological speech processing. It focuses principally on the speech correction and the assistance to learners of Arabic vocabulary. For this purpose, we follow five main phases. The first one is dedicated to evaluate the produced speech by assigning a pronunciation level for each speaker according to their forced alignment score. The second step consists in classifying the Arabic produced speech into healthy or pathological based on two different models: a prosodic modeling based on elocution speed and a phonetic modeling based on comparing between a referenced Probabilistic-Phonetic Model and a speaker model. Third, we localize for each speech sequence classified as pathological the problematic phonemes that degrade pronunciation. We differentiate also two factors which can falsify produced acoustic signals: degraded speech can be generated from pathological problems, or it can be produced by non arabophone pronouncers. Hence, we focus on forced alignment scores. Fourth, we develop a new algorithm to correct pathological pronunciation. We opt of two different solutions: lexical and phonetic. The last task is the conception of an application assisting learners of Arabic vocabulary to improve their pronunciation. The achieved results are encouraging. Moreover, the evaluation and classification of produced acoustic signals are satisfactory, learners of Arabic vocabulary have presented good amelioration using the developed application. A lot of applications that design systems of voice signal processing and platforms of e-learning can enjoy from our proposition.