• 한국음향학회지
• /
• 제28권8호
• /
• pp.832-842
• /
• 2009

Automatic speech recognition (ASR) has been successfully applied to many real human computer interaction (HCI) applications; however, its performance tends to be significantly decreased under noisy environments. The invention of audio visual speech recognition (AVSR) using an acoustic signal and lip motion has recently attracted more attention due to its noise-robustness characteristic. In this paper, we describe our novel integration scheme for AVSR based on a late integration approach. Firstly, we introduce the robust reliability measurement for audio and visual modalities using model based information and signal based information. The model based sources measure the confusability of vocabulary while the signal is used to estimate the noise level. Secondly, the output probabilities of audio and visual speech recognizers are normalized respectively before applying the final integration step using normalized output space and estimated weights. We evaluate the performance of our proposed method via Korean isolated word recognition system. The experimental results demonstrate the effectiveness and feasibility of our proposed system compared to the conventional systems.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2005년도 추계 학술대회 발표논문집
• /
• pp.147-150
• /
• 2005

The performance of speech recognition in car environment is severely degraded when there is music or news coming from a radio or a CD player. Since reference signals are available from the audio unit in the car, it is possible to remove them with an adaptive filter. In this paper, we present experimental results of speech recognition in car environment using the echo canceller. For this, we generate test speech signals by adding music or news to the car noisy speech from Aurora2 DB. The HTK-based continuous HMT system is constructed for a recognition system. In addition, the MMSE-STSA method is used to the output of the echo canceller to remove the residual noise more.

• 디지털융복합연구
• /
• 제16권11호
• /
• pp.335-340
• /
• 2018

정확한 인식률을 보이고 있는 상업적인 음성인식 시스템은 화자종속 고립데이터로부터 학습 모델을 사용한다. 그러나 잡음 환경에서 데이터양에 따라 음성인식의 성능이 저하되는 문제점이 있다. 본 논문에서는 가우시안 분포에서 Maximum Log Likelihood를 이용한 벡터 양자화 기반 음성 인식 성능 향상을 제안한다. 제안하는 방법은 음성에 대한 특징을 가지고 벡터 양자화와 Maximum Log Likelihood 음성 특징 추출 방법을 이용하여 유사 음성에 대한 음성 인식의 정확성을 높이는 최적 학습 모델 구성 방법이다. 이를 위해 HMM을 기반으로 음성 특징을 추출하는 방법을 사용한다. 제안하는 방법을 사용하여 기존 시스템에서 생성되어 사용되는 음성 모델에 대한 부정확한 음성 모델에 대한 정확성을 향상시킬 수 있으므로 음성 인식에 강인한 모델을 구성할 수 있다. 제안하는 방법은 음성 인식 시스템에서 향상된 인식의 정확도를 보인다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2007년도 하계종합학술대회 논문집
• /
• pp.289-290
• /
• 2007

In order to improve the performance of a speech recognition system in the presence of noise, we propose a noise robust front-end using two-channel speech signals by separating speech from noise based on the computational auditory scene analysis (CASA). The main cues for the separation are interaural time difference (ITD) and interaural level difference (ILD) between two-channel signal. As a result, we can extract 39 cepstral coefficients are extracted from separated speech components. It is shown from speech recognition experiments that proposed front-end has outperforms the ETSI front-end with single-channel speech.

• 대한음성학회지:말소리
• /
• 제55권
• /
• pp.131-142
• /
• 2005

A mismatch between the training and the test conditions often causes a drastic decrease in the performance of the speech recognition systems. In this paper, non-linear transformation techniques based on histogram equalization in the acoustic feature space are studied for reducing the mismatched condition. The purpose of histogram equalization(HEQ) is to convert the probability distribution of test speech into the probability distribution of training speech. While conventional histogram equalization methods consider only the probability distribution of a test speech, for noise-corrupted test speech, its probability distribution is also distorted. The transformation function obtained by this distorted probability distribution maybe bring about miss-transformation of feature vectors, and this causes the performance of histogram equalization to decrease. Therefore, this paper proposes a new method of calculating noise-removed probability distribution by using assumption that the CDF of noisy speech feature vectors consists of component of speech feature vectors and component of noise feature vectors, and this compensated probability distribution is used in HEQ process. In the AURORA-2 framework, the proposed method reduced the error rate by over $44\%$ in clean training condition compared to the baseline system. For multi training condition, the proposed methods are also better than the baseline system.

• The Journal of the Acoustical Society of Korea
• /
• 제17권3E호
• /
• pp.3-11
• /
• 1998

Zero-crossings with Peak amplitudes (ZCPA) model motivated by human auditory periphery was proposed to extract reliable features form speech signals even in noisy environments for robust speech recognition. In this paper, the performance of the ZCPA model is further improved by incorporating conventional speech processing techniques into the model output. Spectral and cepstral representations of the ZCPA model output are compared, and the incorporation of dynamic features with several different lengths of time-derivative window are evaluated. Also, comparative evaluations with other front-ends in real-world noisy environments are performed, and result in the superiority of the ZCPA model.

• 한국지능시스템학회논문지
• /
• 제15권6호
• /
• pp.655-660
• /
• 2005

본 논문에서는 인간의 감정 변화에 강인한 음성 인식 기술 개발을 목표로 하여 감정 변화의 영향을 적게 받는 음성 인식시스템의 특징 파라메터에 관한 연구를 수행하였다. 이를 위하여 우선 다양한 감정이 포함된 음성 데이터베이스를 사용하여 감정 변화가 음성 인식 시스템의 성능에 미치는 영향에 관한 연구와 감정 변화의 영향을 적게 받는 음성 인식 시스템의 특징 파라메터에 관한 연구를 수행하였다. 본 연구에서는 LPC 켑스트럼 계수, 멜 켑스트럼 계수, 루트 켑스트럼 계수, PLP 계수와 RASTA 처리를 한 멜 켑스트럼 계수와 음성의 에너지를 사용하였다 또한 음성에 포함된 편의(bias)를 제거하는 방법으로 CMS와 SBR 방법을 사용하여 그 성능을 비교하였다. 실험 결과에서 RASTA 멜 켑스트럼과 델타 켑스트럼을 사용하고 신초편의 제거 방법으로 CMS를 사용한 경우에 HMM 기반의 화자독립 단어 인식기의 오차가 $7.05\%$로 가장 우수한 성능을 나타내었다. 이러한 것은 멜 켑스트럼을 사용한 기준시스템과 비교하여 $59\%$정도 오차가 감소된 것이다.

• 대한음성학회지:말소리
• /
• 제44호
• /
• pp.157-169
• /
• 2002

We propose a weighted filter bank analysis and model adaptation (WFBA-MA) scheme to improve the utilization of uncorrupted or less severely corrupted frequency regions for robust speech recognition. A weighted met frequency cepstral coefficient is obtained by weighting log filter bank energies with reliability coefficients and hidden Markov models are also modified to reflect the local reliabilities. Experimental results on TIDIGITS database corrupted by band-limited noises and car noise indicated that the proposed WFBA-MA scheme utilizes the uncorrupted speech information well, significantly improving recognition performance in comparison to multi-band speech recognition systems.

• 말소리와 음성과학
• /
• 제9권2호
• /
• pp.103-110
• /
• 2017

The pole filtering concept has been successfully applied to cepstral feature normalization techniques for noise-robust speech recognition. In this paper, it is proposed to apply the pole filtering selectively only to the speech intervals, in order to further improve the recognition performance for short utterances in noisy environments. Experimental results on AURORA 2 task with clean-condition training show that the proposed selectively pole-filtered cepstral mean normalization (SPFCMN) and selectively pole-filtered cepstral mean and variance normalization (SPFCMVN) yield error rate reduction of 38.6% and 45.8%, respectively, compared to the baseline system.

• The Journal of the Acoustical Society of Korea
• /
• 제22권3E호
• /
• pp.124-132
• /
• 2003

We propose a new method for speech/non-speech classifiers based on concepts of the adaptive boosting (AdaBoost) algorithm in order to detect speech for robust speech recognition. The method uses a combination of simple base classifiers through the AdaBoost algorithm and a set of optimized speech features combined with spectral subtraction. The key benefits of this method are the simple implementation, low computational complexity and the avoidance of the over-fitting problem. We checked the validity of the method by comparing its performance with the speech/non-speech classifier used in a standard voice activity detector. For speech recognition purpose, additional performance improvements were achieved by the adoption of new features including speech band energies and MFCC-based spectral distortion. For the same false alarm rate, the method reduced 20-50% of miss errors.