• Journal of Korea Multimedia Society
• /
• v.17 no.3
• /
• pp.312-323
• /
• 2014

Voice activity detection is very important process that find voice activity from noisy speech signal for noise cancelling and speech enhancement. Over the past few years, many studies have been made on voice activity detection, it has poor performance for speech signal of sentence form in a low SNR environment. In this paper, it proposed new voice activity detection algorithm that has beginning VAD process using entropy and main VAD process using fuzzy membership shifted c-means clustering. We conduct an experiment in various SNR environment of white noise to evaluate performance of the proposed algorithm and confirmed good performance of the proposed algorithm.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.2
• /
• pp.354-363
• /
• 1994

This paper describes the method that voice traffic is served efficiently in BISDN. Voice is divided into talkspurt and silent period, and it is possible to transmit olny talksurt by the speech activity detection. This paper described the voice traffic control algorithm in the ATM network where cell discarding method is applied to the embedded ADPCM voice data. For traffic control, the cell discarding was used over low priority cells when it overflows the queue threshold. To estimate the efficiency of traffic control algorithm, the computer simuation was performed with cell loss probability, queue length and mean delay as performance parameters. The embedded ADPCM voice coding and cell disscarding resulted in improving the voice cell traffic efficiency and the dynamic control over network congestion.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.6
• /
• pp.759-763
• /
• 2004

In this paper, real-time voice change method using pitch change technique is proposed to change one's voice to the other voice. For this purpose, sampling rate change method using DFT (Discrete Fourier Transform) method and time scale modification method using SOLA (Synchronized Overlap and Add) method is combined to change pitch. In order to evaluate the performance of the proposed method, voice transformation experiments were conducted. Experimental results showed that original speech signal is changed to the other speech signal in which original speaker's identity is difficult to find. The system is implemented using TI TMS320C6711DSK board to verify the system runs in real time.

• Journal of Korea Multimedia Society
• /
• v.16 no.9
• /
• pp.1005-1017
• /
• 2013

Voice activity detection is very important process that voice activity separated form noisy speech signal for speech enhance. Over the past few years, many studies have been made on voice activity detection, but it has poor performance in low signal to noise ratio environment or fickle noise such as car noise. In this paper, it proposed new voice activity detection algorithm using ensemble variance based on wavelet band entropy and soft thresholding method. We conduct a survey in a lot of signal to noise ratio environment of car noise to evaluate performance of the proposed algorithm and confirmed performance of the proposed algorithm.

• Annual Conference on Human and Language Technology
• /
• 2015.10a
• /
• pp.129-132
• /
• 2015

본 논문에서는 단어열 패턴과 리커런트 신경망을 이용한 하이브리드 음성 인식 오류 수정 방법을 제안한다. 음성 인식 결과 문장에서 음성 인식 오류 단어가 발견되었을 경우에 첫째로 단어열 패턴과 그 패턴의 발음열 점수를 통해 1차적 수정을 하고 적절한 패턴을 찾지 못하였을 경우 음절단위로 구성된 Recurrent Neural Network를 통해 단어를 음절단위로 생성하여 2차적으로 오류를 수정한다. 해당 방법론을 한국어로 된 음성 인식 오류와 그 정답 문장으로 구성된 TV 가이드 영역 말뭉치를 바탕으로 성능을 평가하였고, 기존의 단순 단어열 패턴 기반의 음성 인식 오류 수정보다 성능이 향상되었음을 볼 수 있었다. 이 방법론은 음성 인식 오류와 정답의 말뭉치가 필요 없이 옳은 문장으로만 구성된 일반 말뭉치만으로 훈련이 가능하여, 음성 인식 엔진에 의존적이지 않는 강점이 있다.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.2
• /
• pp.75-79
• /
• 2008

In this paper, we propose a novel speech enhancement technique using global soft decision (GSD) based on the probabilistic outputs of support vector machine (SVM). Generally, speech enhancement algorithms applied soft decision gain modification and noise power estimation have bettor performance than those employing hard decision. Especially, global speech absence probability (GSAP), which is known as an effective measure of the speech absence in each frame, has been adopted to SD-based speech enhancement methods. For this reason, we introduce a new GSAP estimated from the probabilistic output of SVM using sigmoid function. The performance of the proposed algorithm is evaluated by the PESQ and MOS test under various noise environments and yields better results compared with the conventional GSD scheme.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.279-282
• /
• 2001

해마다 실시되고 있는 대학수학능력시험의 듣기 평가시험에 있어서, 고사장 및 지점별 음향 성능에 기인한 레벨 편차와 명료도를 산출하여 학생 선발의 공공성과 객관성을 검토해 보고자 한다. 일반적으로 듣기 평가가 이루어지는 각 고사장은 듣기 평가 실시에 지장이 없는 고등학교 교실로 지정하고 있지만, 균등한 음 환경을 제공해야 하는 시험의 성격에 반해 학교 자체의 방송 시설을 그대로 사용하는 것은 평가의 형평성 및 객관성에 충실하지 못하게 되는 요인이 된다. 따라서, 각 고사장의 확성 시스템에 따른 음성 전달품질과 수험생간의 좌석별 음압 레벨 및 명료도를 평가하였고, 실험 결과 RASTI를 비롯한 음성 및 음절 명료도를 나타내는 파라메터에서 좌석별로 큰 편차가 있는 것으로 분석되었다.

• KAPE Magazine
• /
• s.234
• /
• pp.9-11
• /
• 2016

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.7
• /
• pp.668-673
• /
• 2009

In this paper, we propose the minima controlled speech presence uncertainty tracking method to improve a speech enhancement. In the conventional tracking speech presence uncertainty, we propose a method for estimating distinct values of the a priori speech absence probability for different frames and channels. This estimation is inherently based on a posteriori SNR and used in estimating the speech absence probability (SAP). In this paper, we propose a novel estimation of distinct values of the a priori speech absence probability, which is based on minima controlled speech presence uncertainty tracking method, for different frames and channels. Subsequently, estimation is applied to the calculation of speech absence probability for speech enhancement. Performance of the proposed enhancement algorithm is evaluated by ITU-T P. 862 perceptual evaluation of speech quality (PESQ) under various noise environments. We show that the proposed algorithm yields better results compared to the conventional tracking speech presence uncertainty.

• Journal of the Institute of Convergence Signal Processing
• /
• v.4 no.4
• /
• pp.30-39
• /
• 2003

In this paper, we carried out the performance evaluation of HM-Net(Hidden Markov Network) speech recognition system for Korean speech databases. We adopted to construct acoustic models using the HM-Nets modified by HMMs(Hidden Markov Models), which are widely used as the statistical modeling methods. HM-Nets are carried out the state splitting for contextual and temporal domain by PDT-SSS(Phonetic Decision Tree-based Successive State Splitting) algorithm, which is modified the original SSS algorithm. Especially it adopted the phonetic decision tree to effectively express the context information not appear in training speech data on contextual domain state splitting. In case of temporal domain state splitting, to effectively represent information of each phoneme maintenance in the state splitting is carried out, and then the optimal model network of triphone types are constructed by in the parameter. Speech recognition was performed using the one-pass Viterbi beam search algorithm with phone-pair/word-pair grammar for phoneme/word recognition, respectively and using the multi-pass search algorithm with n-gram language models for sentence recognition. The tree-structured lexicon was used in order to decrease the number of nodes by sharing the same prefixes among words. In this paper, the performance evaluation of HM-Net speech recognition system is carried out for various recognition conditions. Through the experiments, we verified that it has very superior recognition performance compared with the previous introduced recognition system.