Voice Signal Research Articles

Product Label, Obstacle and Sign Boards Detection for Visually Impaired People

From the ancient days, visually impaired people had been facing many problems.In this scenario due to heavy traffic there is a chance for them to meet with an accident due to the absence of vision. In addition it becomes necessary for them to depend on others while crossing the roads, to recognize the product labels and to identify the directions in the one way roads. This work helps such people to overcome those problems.This work uses camera module for identifying directions and product labels. The camera captures the label image of the product which is in front of it, detects the product label image and then the image is separated using openCV library. It also uses ultrasonic sensors for identifying obstacles. The output can be heard by impaired people in the form of voice signals through earphones. All this work is carried out by using raspberry pi 3 module B+ with python software. This work helps the visually impaired people to find obstacles which appear in front of them. It also helps in identifying product label and sign boards on the road.

Investigation Results on Individual Differences Appearing in Arousal Evaluation Characteristics Defined on Uttered Voice Signal

Investigation Results on Individual Differences Appearing in Arousal Evaluation Characteristics Defined on Uttered Voice Signal

음성신호 분석을 이용한 EMS 저주파 마사지기가 인체에 미치는 영향 분석

신장은 30%의 기능만 유지되어도 정상적인 삶을 지낼 수 있다. 하지만 이상 증세를 느껴 병 원을 가면 이미 기능이 대부분 상실되어 투석이나 신장을 제거를 해야 하는 상황이 생긴다. 따라서 조 기에 발견하는 것이 중요하다. 신장의 이상 증상으로는 여러 가지가 있지만 그중에 허리의 통증이 있 다. 가정에서도 손쉽게 EMS 저주파 마사지기를 사용하고 있다. 통증 치료를 병원에서 하는 물리치료 가 아니라 간편히 할 수 있는 EMS 저주파 마사지기를 사용한 후, 음성 녹음을 통해 음성 분석 프로 그램 Praat을 이용하여 음성신호 분석을 통해 EMS 저주파 마사지기가 인체에 미치는 영향에 대해 분석하였다. 이를 위해, 허리 통증을 느끼는 사람을 대상으로 EMS 저주파 마사지 치료 전과 후의 음 성을 수집하고 Pitch 주파수 및 제1포먼트 주파수 대역폭의 변화를 분석하였다.

The voice tone frequency estimate of the user authentication system

The directions for improving the quality characteristics of voice authentication systems in different access systems are analyzed and investigated in the article. One of the main areas of enhancing the quality characteristics of these user authentication systems is related to the use of phase information of the voice signal. The object of the study is the process of digital signal processing in voice authentication systems. The relevant scientific task of investigating new procedures to refine estimates of pitch frequency obtained on the basis of the analysis of the amplitude-frequency spectrum is solved. The estimates were refined based on the use of phase data of a voice signal, as well as the estimation of the pitch frequency in the process of obtaining kepstral coefficients. The purpose of this work is to investigate the effect of phase data on the accuracy of estimating the pitch frequency in the process of forming a user pattern. The results have been obtained in the process of statistical analysis of the simulation results using model and experimental voice data of the user authentication system. The phase data of the voice signal allow to obtain adequate and reliable estimates in the process of spectral analysis. However, in the presence of errors associated with gross errors, such as the acceptance of the pitch of the frequencies of the first or second formants, preference should be given to the estimate obtained in the process of calculating kepstral coefficients. The influence of sample length and their overlap in the process of calculating kepstral coefficients is investigated. The presented research results should be used in voice authentication systems, improvement of speech recognition systems, as well as in solving problems of speaker identification.

Direct-Mapping-Based MIMO-FBMC Underwater Acoustic Communication Architecture for Multimedia Signals

In this paper, a direct-mapping (DM)-based multi-input multi-output (MIMO) filter bank multi-carrier (FBMC) underwater acoustic multimedia communication architecture (UAMCA) is proposed. The proposed DM-based MIMO-FBMC UAMCA is rare and non-obvious in the underwater multimedia communication research topic. The following are integrated into the proposed UAMCA: A 2 × 2 DM transmission mechanism, a (2000, 1000) low-density parity-check code encoder, a power assignment mechanism, an object-composition petrinet mechanism, adaptive binary phase shift keying modulation and 4-offset quadrature amplitude modulation methods. The multimedia signals include voice, image, and data. The DM transmission mechanism in different spatial hardware devices transmits different multimedia packets. The proposed underwater multimedia transmission power allocation algorithm (UMTPAA) is simple, fast, and easy to implement, and the threshold transmission bit error rates (BERs) and real-time requirements for voice, image, and data signals can be achieved using the proposed UMTPAA. The BERs of the multimedia signals, data symbol error rates of the data signals, power saving ratios of the voice, image and data signals, mean square errors of the voice signals, and peak signal-to-noise ratios of the image signals, for the proposed UAMCA with a perfect channel estimation, and channel estimation errors of 5%, 10%, and 20%, respectively, were explored and demonstrated. Simulation results demonstrate that the proposed 2 × 2 DM-based MIMO-FBMC UAMCA is suitable for low power and high speed underwater multimedia sensor networks.

Spreading-Based Voice Encryption by Means of OVSF Codes

This paper presents a new methodology to encrypt voice signals, in such a way that they simulate being a noise signal. The objective is to obtain a signal that does not generate suspicions about its content, while protecting the message. The process is based on the spread and scrambling of the signal through the use of OVSF (orthogonal variable spreading factor) codes. The security of the method is based mainly on the input value used for the randomization phase. From a computational cost point of view, the use of fixed-length codes allows for a shorter retrieval time compared to a similar method based on variable-length codes. Regarding the encrypted signal, its main characteristic is its high entropy (very close to the ideal value). Finally, in the recovery process, a signal identical to the original is obtained.

Investigating voice as a biomarker: Deep phenotyping methods for early detection of Parkinson's disease.

Voice technology has grown tremendously in recent years and using voice as a biomarker has also been gaining evidence. We demonstrate the potential of voice in serving as a deep phenotype for Parkinson's Disease (PD), the second most common neurodegenerative disorder worldwide, by presenting methodology for voice signal processing for clinical analysis. Detection of PD symptoms typically requires an exam by a movement disorder specialist and can be hard to access and inconsistent in findings. A vocal digital biomarker could supplement the cumbersome existing manual exam by detecting and quantifying symptoms to guide treatment. Specifically, vocal biomarkers of PD are a potentially effective method of assessing symptoms and severity in daily life, which is the focus of the current research. We analyzed a database of PD patient and non-PD subjects containing voice recordings that were used to extract paralinguistic features, which served as inputs to machine learning models to predict PD severity. The results are presented here and the limitations are discussed given the nature of the recordings. We note that our methodology only advances biomarker research and is not cleared for clinical use. Specifically, we demonstrate that conventional machine learning models applied to voice signals can be used to differentiate participants with PD who exhibit little to no symptoms from healthy controls. This work highlights the potential of voice to be used for early detection of PD and indicates that voice may serve as a deep phenotype for PD, enabling precision medicine by improving the speed, accuracy, accessibility, and cost of PD management.

Gender classification based on voice signals using fuzzy models and optimization algorithms

This paper describes a gender classification scheme based on voice signals in which 16 different fuzzy models are proposed and optimized using four bio-inspired optimization algorithms and the quasi-Newton method. The classification scheme considers four data sets and five different voice features to define the input values of an algorithm in the optimization process. The inputs of each fuzzy model define the mean and variance of their Gaussian membership functions, and their fitness is evaluated by the input values of the algorithm and mean squared error as objective function to be minimized. A comparative analysis between models, algorithms and data sets is made to obtain conclusions according to the results of each optimized model.

Smart Self Defense Security System for Women

Women all over the world are undergoing and are subjected to immoral physical persecution. Security for women is always a major problem as the crimes and persecution over women and girls is increasing day-by-day. They all need is a gadget that would be transmit easily. This paper presents a self-defense security using smart phone and an embedded unit. This idea is designed using ATmega2560 platform. The embedded unit and the smart phone are linked by Bluetooth paring. On receiving the voice command by the developed app in the smart phone, the GPS finds the person by locating out of latitude and longitude while GSM sends this information and the present status of the person whose voice has been received to the pre-specified mobile number and to the nearby police station. The voice command will lead to the paring of the Bluetooth. One of the main function done by this security system is the self-defense. This self-defense is generated by the electric shock providing circuit. The electric shock providing circuit is a circuit which operates on the principle of a mosquito bat. The shock providing circuit is switched ON and OFF based on the received voice command. If the voice signal is high enough a moderate amount of voltage is given out, if low vice versa.

Design and analysis of radio antenna monopole arrays in the military world

The analysis and design of the proposed antennas is designated for receiving voice signals from Radio in general are out-of-prominent antennas and smaller indoor antennas, in this research, the authors aim to design and analyse an antenna with a simple design, a small dimension efficient, very easy installation, cheap and meet the antenna characteristics to support the movement in combat training within the Army. The antenna to be discussed is a monopole antenna Arrays with a minimalist shape that works at frequencies between 350MHz - 480MHz by adjusting the length of the horizontal elemental Active distance between the outer and inner to get the impedance matching. In order to obtain a unidirectional pattern, a metallic plane reflector is placed at a distance of about λ/4 from the array board: predicted gain (11 dB for 5 × 1 array and 15 dB for 9 × 1 array) and bandwidth (around 600 MHz) are larger than for conventional micro strip patch arrays having similar board dimensions. the proposed antenna has been numerically analysed for wireless communication systems.

Patterned Endorsement Framework for Defense Sector using Two Level Voice Authentication Process

The guard segment considers security usage as the essential and critical errand to be executed so usage ought to be made with more concern. One of the ideal arrangements is Voice for upgrading the security framework in which the confirmation of clients is performed with the guide of their voice signals. This confirmation has been performed by actualizing Security Voice Authentication System (SVAS) strategy in our past work. Be that as it may, this philosophy isn't a lot of viable on the insider aggressors. The security framework would be separated by the vindictive/narrow-minded clients as they go about as genuine clients by transmitting the voice sign of unique clients. This deficiency can be overwhelmed by presenting a procedure called Two Level Voice Authentication System (TLVAS). This examination includes producing mystery keys at first by methods for arbitrary qualities inputted by the client. It will give security in the two-level procedure which incorporates the respectability of the handling ought to be solid. The sign must be changed over which control the figuring techniques. This strategy is seen as increasingly powerful against the insider assailants by avoiding the Compromisation created by them. Installing procedure used to execute for doing the exhibition investigation and the examination is completed by methods for expanded security level.

Patterned Endorsement Framework for Defense Sector using Two Level Voice Authentication Process

The guard segment considers security usage as the essential and critical errand to be executed so usage ought to be made with more concern. One of the ideal arrangements is Voice for upgrading the security framework in which the confirmation of clients is performed with the guide of their voice signals. This confirmation has been performed by actualizing Security Voice Authentication System (SVAS) strategy in our past work. Be that as it may, this philosophy isn't a lot of viable on the insider aggressors. The security framework would be separated by the vindictive/narrow-minded clients as they go about as genuine clients by transmitting the voice sign of unique clients. This deficiency can be overwhelmed by presenting a procedure called Two Level Voice Authentication System (TLVAS). This examination includes producing mystery keys at first by methods for arbitrary qualities inputted by the client. It will give security in the two-level procedure which incorporates the respectability of the handling ought to be solid. The sign must be changed over which control the figuring techniques. This strategy is seen as increasingly powerful against the insider assailants by avoiding the Compromisation created by them. Installing procedure used to execute for doing the exhibition investigation and the examination is completed by methods for expanded security level.

Statistical Analysis of Lower and Raised Pitch Voice Signal and Its Efficiency Calculation

Statistical Analysis of Lower and Raised Pitch Voice Signal and Its Efficiency Calculation

Voice pattern recognition using Mel-Frequency Cepstral Coefficient and Hidden Markov Model for bahasa Madura

Voice recognition is one part of an application that allows a device to recognize spoken words by digitizing words and matching digital signals with a particular pattern stored in a device. Spoken words are converted into digital signals by converting voice waves into a set of numbers which is then compared with the voice pattern to identify the words. MFCC can be an alternative method to solve the problem of voice extraction because this method is reliable for recognizing the unique features of human voice. Hidden Markov Model is used to recognize the voice pattern, so it can be used to compare the voice signal obtained from e-learning with the trained voice signal. Bahasa Madura is a regional language used by ethnic Madurese to communicate daily. Currently the number of Madurese people who understand this language is reduced so that the use of Bahasa Madura is also reduced. Therefore, it is necessary to conduct speech recognition research in Madura Language as one of effort to preserve and develop the use of Regional Language. The experimental results show that the average accuracy for testing the system with one model is 85% and the average accuracy for testing the system with multi model is 90%.

Voice compression using discrete cosine transform and wavelet transform

Compressing information, such as voice, allows freeing up memory space, better utilizing bandwidth, in other words taking advantage of resources more efficiently in a time where technologies such as Voice over Internet Protocol are booming and the demand for information it requires higher speeds and quality. Therefore, this paper presents the acquisition and digital processing of voice signals, as well as the application of the discrete cosine transform and the wavelet transform using Matlab software version 2017b, licensed by the Technological University of Bolivar. Likewise, a criterion is established in order to eliminate little relevant information without affecting the quality of the output signal for which indicators such as the signal to noise ratio and the mean square error are calculated.

Variable Step-size LMS Algorithm Based on Discrete Cosine Transform for Voice Signal Denoising

Starting from the characteristic of voice signal itself, a new variable step-size LMS algorithm based on DCT is proposed in order to improve the comprehensive performance in adaptive noise cancellation, which combines the merits of normalized DCT-LMS algorithm and variable step-size LMS algorithm, and give full play to the decorrelation capability of DCT and rapid convergence effect of variable step-size algorithm. The c simulation results show that the new algorithm has faster convergence rate and smaller steady state error compared with traditional LMS and NLMS algorithm, however, the computational complexity is comparable to NLMS algorithm at the same time.

Research on Voice Print Recognition Based on Wavelet Analysis and BP-GA

Aiming at the problems of common algorithm in voice print recognition, to improve the performance of recognition system, this paper put forward a new recognition algorithm based on wavelet analysis and BP-GA optimization algorithm. The algorithm, firstly extracts time-domain and frequency-domain characteristic variation of voice signal using wavelet transform, and then trains the variation data with BP-GA algorithm, which overcomes the problem of traditional multi-layers neural network’s local minimum value, lastly puts the trained data into a designed network to form the final voice print algorithm. Simulation experiment showed compared with current recognition algorithm, the algorithm in this paper has the advantage of fast recognition velocity, high recognition rate, low fault rate, automatic error correcting, and strong robustness with different voice-makers.

Generation of Telugu Emotional Speech from Neutral Speech by using Voice Morphing Algorithm

Voice morphing is the conversion of one voice signal into another voice signal i.e. source voice signal into target voice signal. Like image morphing, the aim of speech morphing is to preserve the pooled properties of the source signal and target signal during the generation of a smooth conversion between them. The major properties of speech signal are its pitch and envelope information. Pitch, duration and intensity of speech signal parameters are considered to get emotional voice from neutral or emotion. We are using TD-PSOLA technique to extract the prosodic parameters and during the training phase.

A new method for voice signal features creation

Digital audio is one of the most important types of data at present. It is used in several applications, such as human knowledge and many security and banking applications. A digital voice signal is usually of a large size where the acoustic signal consists of a set of values distributed in one column (one channel) (mono signal) or distributed in two columns (two channels) (stereo signal), these values usually are the results of sampling and quantization of the original analogue voice signal. In this paper we will introduce a method which can be used to create a signature or key, which can be used later to identify or recognize the wave file. The proposed method will be implemented and tested to show the accuracy and flexibility of this method.

NIML: non‐intrusive machine learning‐based speech quality prediction on VoIP networks

Voice over Internet Protocol (VoIP) networks have recently emerged as a promising telecommunication medium for transmitting voice signal. One of the essential aspects that interests researchers is how to estimate the quality of transmitted voice over VoIP for several purposes such as design and technical issues. Two methodologies are used to evaluate the voice, which are subjective and objective methods. In this study, the authors propose a non-intrusive machine learning-based (NIML) objective method to estimate the quality of voice. In particular, they build a training set of parameters – from the network and the voice itself – along with the quality of voices as labels. The voice quality is estimated using the perceptual evaluation of speech quality (PESQ) method as an intrusive algorithm. Then, the authors use a set of classifiers to build models for estimating the quality of the transmitted voice from the training set. The experimental results show that the classifier models have a valuable performance where Random Forest model has superior results compared to other models of precision 94.1%, recall 94.2%, and receiver operating characteristic area 99.2% as evaluation metrics.