检索结果-内蒙古大学图书馆

IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhen, Kai Lee, Mi Suk Sung, Jongmo Beack, Seungkwon Kim, Minje Indiana Univ Luddy Sch Informat Comp & Engn Bloomington IN 47405 USA Indiana Univ Cognit Sci Program Bloomington IN USA Elect & Telecommun Res Inst Daejeon South Korea

ISBN: (纸本)9781509066315

Scalability and efficiency are desired in neural speech codecs, which supports a wide range of bitrates for applications on various devices. We propose a collaborative quantization (CQ) scheme to jointly learn the codebook of LPC coefficients and the corresponding residuals. CQ does not simply shoehorn LPC to a neural network, but bridges the computational capacity of advanced neural network models and traditional, yet efficient and domain-specific digital signal processing methods in an integrated manner. We demonstrate that CQ achieves much higher quality than its predecessor at 9 kbps with even lower model complexity. We also show that CQ can scale up to 24 kbps where it outperforms AMR-WB and Opus. As a neural waveform codec, CQ models are with less than 1 million parameters, significantly less than many other generative models.

关键词： Speech coding linear predictive coding deep neural network residual learning model complexity

来源：评论

学校读者我要写书评

暂无评论

An Acoustic Signal Based Language Independent Lip Synchronization Method and Its Implementation via Extended LPC 28

An Acoustic Signal Based Language Independent Lip Synchroniz...

引用

28th Signal Processing and Communications Applications Conference (SIU)

作者： Cankurtaran, Halil Said Boyaci, Ali Yarkan, Serhan Istanbul Ticaret Univ Elekt Elekt Muhendisligi Bolumu Kucukyali E5 KavsagiInonu Cad4 Kucukyali TR-34840 Istanbul Turkey Istanbul Ticaret Univ Bilgisayar Muhendisligi Bolumu Kucukyali E5 KavsagiInonu Cad4 Kucukyali TR-34840 Istanbul Turkey

ISBN: (纸本)9781728172064

Processing human speech with the use of digital technologies leads to several important fields of research. Speech-to-text and lip-syncing are among the instances of relevant prominent research areas. In this regard, audio-visualization of acoustic signals, providing visual aid in real-time for disabled people, and realization of text-free animation applications are just to name a few. Therefore, in this study, a language-independent lip-sync method that is based on extended linear predictive coding is proposed. The proposed method operates on baseband electrical signal that is acquired by a standard single-channel off-the-shelf microphone and exploits the statistical characteristics of acoustic signals produced by human speech. In addition, the proposed method is implemented on an embedded system, tested, and its performance is evaluated. Results are given along with discussions and future directions.

关键词： formant frequency linear predictive coding lip sync

来源：评论

学校读者我要写书评

暂无评论

Convolutional Neural Network applied in EEG imagined phoneme recognition system

Convolutional Neural Network applied in EEG imagined phoneme...

引用

International Symposium on Advanced Topics in Electrical Engineering (ATEE)

作者： Ana-Luiza Rusnac Ovidiu Grigore Faculty of Electronics Telecommunications and Information Technology Univeristy “Politehnica” of Bucharest Bucharest Romania

ISBN: (纸本)9781665418782;9781665448079

Speech is a skill that most of the time we take for granted. In reality, this ability is a complex mechanism which requires thoughts to be translated into words who are further transposed into sounds, a mechanism which involves precise coordination of several muscles and joints. In some cases, this complex mechanism can no longer be performed and may be accompanied by almost complete loss of motor activity such in diseases as: stroke, Lock-Down syndrome, amyotrophic lateral sclerosis, cerebral palsy etc. The most recent method that aims to supplement the speech mechanism is imaginary speech recognition using electroencephalographic (EEG) signals, by using complex computing mechanisms like Deep Learning (DL) in order to decode the thoughts. In this paper we aim to recognize three types of clustered phonemes using conventional speech recognition techniques, like Mel-Cepstral Coefficients (MFCC) and linear predictive coding (LPC) combined with Convolutional Neural Networks (CNN). We compared four types of features extraction: MFCC, LPC, MFCC+ LPC combined into 1-channel matrix and MFCC+ LPC combined into a 2-channel matrix. We showed that MFCC coefficients offer a better accuracy than LPC and that concatenating MFCC and LPC into a 2-channel matrix we obtain a better performance than combining them into 1-channel matrix.

关键词： Support vector machines Image recognition Neural networks Muscles Feature extraction Electroencephalography linear predictive coding

来源：评论

学校读者我要写书评

暂无评论

Feature Extraction Methods based on linear predictive coding and Wavelet Packet Decomposition for Recognizing Spoken Words in Malayalam

Feature Extraction Methods based on Linear Predictive Coding...

引用

2nd International Conference on Advances in Computing and Communications (ICACC)

作者： Sunny, Sonia Peter, David S. Jacob, K. Poulose Cochin Univ Sci & Technol Dept Comp Sci Kochi 682022 Kerala India Cochin Univ Sci & Technol Sch Engn Kochi 682022 Kerala India

ISBN: (纸本)9780769547237;9781467319119

Speech signals are one of the most important means of communication among the human beings. In this paper, a comparative study of two feature extraction techniques are carried out for recognizing speaker independent spoken isolated words. First one is a hybrid approach with linear predictive coding (LPC) and Artificial Neural Networks (ANN) and the second method uses a combination of Wavelet Packet Decomposition (WPD) and Artificial Neural Networks. Voice signals are sampled directly from the microphone and then they are processed using these two techniques for extracting the features. Words from Malayalam, one of the four major Dravidian languages of southern India are chosen for recognition. Training, testing and pattern recognition are performed using Artificial Neural Networks. Back propagation method is used to train the ANN. The proposed method is implemented for 50 speakers uttering 20 isolated words each. Both the methods produce good recognition accuracy. But Wavelet Packet Decomposition is found to be more suitable for recognizing speech because of its multi-resolution characteristics and efficient time frequency localizations.

关键词： Speech Recognition Feature Extraction linear predictive coding Wavelet Packet Decomposition Neural Networks

来源：评论

学校读者我要写书评

暂无评论

An integrated approach for identification of exon locations using recursive Gauss Newton tuned adaptive Kaiser window

引用

GENOMICS 2019年第3期111卷 284-296页

作者： Das, Lopamudra Nanda, Sarita Das, J. K. KIIT Univ Sch Elect Engn Bhubaneswar Odisha India

Identification of exon location in a DNA sequence has been considered as the most demanding and challenging research topic in the field of Bioinformatics. This work proposes a robust approach combining the Trigonometric mapping with Adaptive tuned Kaiser Windowing approach for locating the protein coding regions (EXONS) in a genetic sequence. For better convergence as well as improved accurateness, the side lobe height control parameter (beta) of Kaiser Window in the proposed algorithm is made adaptive to track the changing dynamics of the genetic sequence. This yields better tracking potential of the anticipated Adaptive Kaiser algorithm as it uses the recursive Gauss Newton tuning which in turn utilizes the covariance of the error signal to tune the beta factor which has been shown through numerous simulation results under a variety of practical test conditions. A detailed comparative analysis with the existing mapping schemes, windowing techniques, and other signal processing methods like SVD, AN, DFT, STDFT, WT, and ST has also been included in the paper to focus on the strength and efficiency of the proposed approach. Moreover, some critical performance parameters have been computed using the proposed approach to investigate the effectiveness and robustness of the algorithm. In addition to this, the proposed approach has also been successfully applied on a number of benchmark gene sets like Musmusculus, Homosapiens, and C. elegans, etc., where the proposed approach revealed efficient prediction of exon location in contrast to the other existing mapping methods.

关键词： Trigonometric mapping linear predictive coding Recursive Gauss Newton tuning Adaptive Kaiser window Exon Receiver operating characteristics

来源：评论

学校读者我要写书评

暂无评论

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

引用

BIO-ALGORITHMS AND MED-SYSTEMS 2019年第1期15卷

作者： Vishwanath, Bhagyalakshmi Pujeri, Ramchandra Vittal Devanagavi, Geeta REVA Univ Bangalore 560064 Karnataka India MIT Coll Engn Pune 411038 Maharashtra India

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 cl

关键词： arrhythmia classification Gabor transform linear predictive coding probabilistic principal component analysis model support vector neural network

来源：评论

学校读者我要写书评

暂无评论

Automated tongue-twister phrase-based screening for Cerebellar Ataxia using Vocal tract Biomarkers 41

Automated tongue-twister phrase-based screening for Cerebell...

引用

41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

作者： Kashyap, Bipasha Pathirana, Pubudu N. Horne, Malcolm Power, Laura Szmulewicz, David Deakin Univ Sch Engn Networked & Sensing Control NSC Lab Waurn Ponds Vic Australia Florey Inst Neurosci & Mental Hlth Parkville Vic Australia Royal Victorian Eye & Ear Hosp RVEEH Balance Disorders & Ataxia Serv St Andrews Pl East Melbourne Vic Australia Alfred Hosp Cerebellar Ataxia Clin Praran Vic Australia

ISBN: (纸本)9781538613115

Cerebellar Ataxia (CA) is a neurological condition that leads to uncoordinated muscle movements, even affecting the production of speech. Effective biomarkers are necessary to produce an objective decision-making support tool for early diagnosis of CA in non-clinical environments. This paper investigates the reliability and effectiveness of vocal tract acoustic biomarkers for assessing CA speech. These features were tested on a database consisting of 52 clinically rated tongue-twister phrase 'British Constitution' and its 4 consonant-vowel (CV) excerpts /ti/, /ti/, /tion/ acquired from 30 ataxic patients and 22 healthy controls. Such a marker could be applied to objectively assess the severity of CA from a simple speaking test, contributing to the possibility of being translated into a computer based automatic module to screen the disease from the speech. All the vocal tract features explored in this study were statistically significant using Kolmogorov-Smirnov test at 5% level in distinguishing healthy and CA speech. Several machine learning classifiers with 5-fold cross-validations were implemented on the vocal features. It was observed that the intensity ratios corresponding to the 4 C-V excerpts in CA group showed an increased variability and produced the best classification accuracy of 84.6% using KNN classifier. Results motivate the use of vocal tract features for monitoring CA speech.

关键词： speech biomarker linear predictive coding spectrogram vocal tract measures acoustic analysis cerebellar ataxia

来源：评论

学校读者我要写书评

暂无评论

Speaker Recognition Based on Long Short-Term Memory Networks

Speaker Recognition Based on Long Short-Term Memory Networks

引用

IEEE International Conference on Signal and Image Processing (ICSIP)

作者： Qihang Xu Mingjiang Wang Changlai Xu Lu Xu Harbin Institute of Technology Shenzhen Shenzhen China

ISBN: (数字)9781728168968

ISBN: (纸本)9781728168975

Speaker recognition is the process of identifying the speaker's identity by extracting the acoustic features in the speaker's audio for recognition. It is mainly the perception and simulation of the speaker's vocal tract information and the human ear's auditory information, which has profound significance in the fields of human daily life and military affairs. In order to solve this problem, this paper proposes a speaker recognition algorithm based on Long Short-Term Memory Networks (LSTM). The feature uses a method that combines linear predictive coding (LPC) with Log-Mel spectrum. The d-vector output through the LSTM network is classified using the Softmax loss function. This method is applied to the VCTK audio data set. The experimental results show that the recognition rate of this method reaches 94.9%.

关键词： Image recognition Conferences Ear Feature extraction Prediction algorithms linear predictive coding Speaker recognition

来源：评论

学校读者我要写书评

暂无评论

An Acoustic Signal Based Language Independent Lip Synchronization Method and Its Implementation via Extended LPC

An Acoustic Signal Based Language Independent Lip Synchroniz...

引用

IEEE Signal Processing and Communications Applications (SIU)

作者： Halil Said Cankurtaran Ali Boyacı Serhan Yarkan İstanbul Ticaret Üniversitesi Elektrik-Elektronik Mühendisligi Bölümü İstanbul Türkiye İstanbul Ticaret Üniversitesi Bilgisayar Mühendisligi Bölümü İstanbul Türkiye

ISBN: (数字)9781728172064

ISBN: (纸本)9781728172071

Processing human speech with the use of digital technologies leads to several important fields of research. Speech- to-text and lip-syncing are among the instances of relevant prominent research areas. In this regard, audio-visualization of acoustic signals, providing visual aid in real-time for disabled people, and realization of text-free animation applications are just to name a few. Therefore, in this study, a language-independent lip-sync method that is based on extended linear predictive coding is proposed. The proposed method operates on baseband electrical signal that is acquired by a standard single-channel off-the-shelf microphone and exploits the statistical characteristics of acoustic signals produced by human speech. In addition, the proposed method is implemented on an embedded system, tested, and its performance is evaluated. Results are given along with discussions and future directions.

关键词： Acoustics Synchronization Lips linear predictive coding Internet Visualization Standards

来源：评论

学校读者我要写书评

暂无评论

Exploring vowel formant estimation through simulation-based techniques

引用

LINGUISTICS VANGUARD 2020年第s1期6卷

作者： Kendall, Tyler Vaughn, Charlotte Univ Oregon Dept Linguist Eugene OR 97403 USA

This paper contributes insight into the sources of variability in vowel formant estimation, a major analytic activity in sociophonetics, by reviewing the outcomes of two simulations that manipulated the settings used for linear predictive coding (LPC)-based vowel formant estimation. Simulation 1 explores the range of frequency differences obtained when minor adjustments are made to LPC settings, and measurement time points around the settings used by trained analysts, in order to determine the range of variability that should be expected in sociophonetic vowel studies. Simulation 2 examines the variability that emerges when LPC settings are varied combinatorially around constant default settings, rather than settings set by trained analysts. The impacts of different LPC settings are discussed as a way of demonstrating the inherent properties of LPC-based formant estimation. This work suggests that differences more fine-grained than about 10 Hz in F1 and 15-20 Hz in F2 are within the range of LPC-based formant estimation variability.

关键词： Sociophonetics vowels formant estimation linear predictive coding methods

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：