Wyniki wyszukiwania - BazTech

1

HMM-based phoneme speech recognition system for the control and command of industrial robots

Naik Adwait

Technical Transactions

|

2021

|

Vol. 118, iss. 1

art. no. e2021002

EN

n recent years, the integration of human-robot interaction with speech recognition has gained a lot of pace in the manufacturing industries. Conventional methods to control the robots include semi-autonomous, fully-autonomous, and wired methods. Operating through a teaching pendant or a joystick is easy to implement but is not effective when the robot is deployed to perform complex repetitive tasks. Speech and touch are natural ways of communicating for humans and speech recognition, being the best option, is a heavily researched technology. In this study, we aim at developing a stable and robust speech recognition system to allow humans to communicate with machines (robotic-arm) in a seamless manner. This paper investigates the potential of the linear predictive coding technique to develop a stable and robust HMM-based phoneme speech recognition system for applications in robotics. Our system is divided into three segments: a microphone array, a voice module, and a robotic arm with three degrees of freedom (DOF). To validate our approach, we performed experiments with simple and complex sentences for various robotic activities such as manipulating a cube and pick and place tasks. Moreover, we also analyzed the test results to rectify problems including accuracy and recognition score.

2

Probabilistic principal component analysis-based dimensionality reduction and optimization for arrhythmia classification using ECG signals

Vishwanath Bhagyalakshmi, Pujeri Ramchandra Vittal, Devanagavi Geeta

Bio-Algorithms and Med-Systems

|

2019

|

Vol. 15, no. 1

art. no. 20180037

EN

Electrocardiogram (ECG) is an electrical signal that contains data about the state and functions of the heart and can be used to diagnose various types of arrhythmias effectively. The modeling and simulation of ECG under different conditions are significant to understand the function of the cardiovascular system and in the diagnosis of heart diseases. Arrhythmia is a severe peril to the patient recovering from acute myocardial infarction. The reliable detection of arrhythmia is a challenge for a cardiovascular diagnostic system. As a result, a considerable amount of research has focused on the development of algorithms for the accurate diagnosis of arrhythmias. In this paper, a system for the classification of arrhythmia is developed by employing the probabilistic principal component analysis (PPCA) model. Initially, the cluster head is selected for the effective transmission of ECG signals of patients using the adaptive fractional artificial bee colony algorithm, and multipath routing for transmission is selected using the fractional bee BAT algorithm. Features such as wavelet features, Gabor transform, empirical mode decomposition, and linear predictive coding features are extracted from the ECG signal with high dimension (which are reduced using PPCA) and finally given to the proposed classifier called adaptive genetic-bat (AGB) support vector neural network (which is trained using the AGB algorithm) for arrhythmia detection. The experimentation of the proposed system is done based on evaluation metrics, such as the number of alive nodes, normalized network energy, goodput, and accuracy. The proposed method obtained a classification accuracy of 0.9865 and a goodput of 0.0590 and provides a better classification of arrhythmia. The experimental results show that the proposed system is useful for the classification of arrhythmias, with a reasonably high accuracy of 0.9865 and a goodput of 0.0590. The validation of the proposed system offers acceptable results for clinical implementation.

3

A method of designing an adaptive uniform quantizer for LPC coefficients quantization

Eskic Z., Peric Z., Nikolic J.

Przegląd Elektrotechniczny

|

2011

|

R. 87, nr 7

245-248

EN

This paper proposes a method of designing the adaptive uniform quantizer for frame by frame LPC coefficients quantization. The method firstly determines the support region thresholds of two uniform quantizers designated to quantize the minimal and the maximal value of LPC coefficients of each frame. Based on this, the uniform quantizer thresholds estimation for LPC coefficients quantization are provided. The results obtained by testing the proposed method in processing the speech signal from the TIMIT data base are presented and disscused in the paper.

PL

W artykule zaproponowano metodę zuniformowane go adaptacyjne kwantowania współczynnika LPC (linear prediction coders). Początkowo obliczana jest minimalna i maksymalna wartość LPC dla każdej ramki. Następnie zuniformowany współczynnik jest określany. Zaprezentowano test metody na przykładzie przetwarzania sygnału mowy z bazy TIMIT.

4

Kompresja stratna dźwięku

Grad L.

Zeszyty Naukowe Warszawskiej Wyższej Szkoły Informatyki

|

2007

|

nr 2

39--58

PL

W artykule przedstawione zostały elementarne wiadomości z zakresu kompresji stratnej dźwięku. Przedstawiony został liniowy model predykcji, wykorzystywany w kompresji dźwięku w paśmie telefonicznym oraz opisane zostały transformacje ortogonalne i model psychoakustyczny człowieka mające podstawowe znaczenia w kompresji dźwięku wysokiej jakości.

EN

This paper presents an overview of the basic information on sound loose compression. Linear predictive coding, which is used in voice compression in the phone band, as well as orthogonal transforms and psychoacoustic model, which are very important in high level sound compression (voice, speech), are revised.This paper presents an overview of the basic information on sound loose compression. Linear predictive coding, which is used in voice compression in the phone band, as well as orthogonal transforms and psychoacoustic model, which are very important in high level sound compression (voice, speech), are revised.

5

Acoustic Properties of Polish Vowels

Izydorczy J., Kłosowski P.

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

1999

|

Vol. 47, nr 1

29-37

EN

The article presents physical basics of human speech production. There are considered the most fundamental phonemes-vowels. Acoustic properties of vowels are extracted. It is observed that the most important are positions and shapes of first two formants. The rest of the spectral attributes of vowels, exception is voiced excitation, can be canelled. Vowels are still recognized by human. In the article acoustic model of vowels production is propsed and tested. Results are presented in graphic manner.