• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.131-134
• /
• 2004

본 논문에서는 GMM(Gaussian Mixture Model)을 이용한 문맥독립 화자 검증 시스템을 구현한 후, arctan 함수를 이용한 정규화 방법을 사용하여 화자검증실험을 수행하였다. 특징파라미터로서는 선형예측방법을 이용한 켑스트럼 계수와 회귀계수를 사용하고 화자의 발성 변이를 고려하여 CMN(Cepstral Mean Normalization)을 적용하였다. 화자모델 생성을 위한 학습단에서는 화자발성의 음향학적 특징을 잘 표현할 수 있는 GMM(Gaussian Mixture Model)을 이용하였고 화자 검증단에서는 ML(Maximum Likelihood)을 이용하여 유사도를 계산하고 기존의 정규화 방법과 arctan 함수를 이용한 방법에 의해 정규화된 점수(score)와 미리 정해진 문턱값과 비교하여 검증하였다. 화자 검증 실험결과, arctan 함수를 부가한 방법이 기존의 방법보다 항상 향상된 EER을 나타냄을 확인할 수 있었다.

• Journal of the Korea Society of Computer and Information
• /
• v.5 no.2
• /
• pp.90-97
• /
• 2000

This paper describes an speaker-adaptive speech recognition system which make a reliable recognition of speech signal for new speakers. In the Proposed method, an speech spectrum of new speaker is adapted to the reference speech spectrum by using Parameters of a 1st linear transformation network at the front of phoneme classification neural network. And the recognition system is based on semicontinuous HMM(hidden markov model) which use the multilayer perceptron as a fuzzy vector quantizer. The experiments on the isolated word recognition are performed to show the recognition rate of the recognition system. In the case of speaker adaptation recognition, the recognition rate show significant improvement for the unadapted recognition system.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.29 no.6
• /
• pp.1393-1401
• /
• 2019

Recently as voice authentication function is installed in the system, it is becoming more important to accurately authenticate speakers. Accordingly, a model for verifying speakers in various ways has been suggested. In this paper, we propose a new method for verifying speaker verification using a Short-time Fourier Transform(STFT). Unlike the existing Mel-Frequency Cepstrum Coefficients(MFCC) extraction method, we used window function with overlap parameter of around 66.1%. In this case, the speech characteristics of the speaker with the temporal characteristics are studied using a deep running model called RNN (Recurrent Neural Network) with LSTM cell. The accuracy of proposed model is around 92.8% and approximately 5.5% higher than that of the existing speaker certification model.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.6
• /
• pp.679-686
• /
• 2020

In this thesis, an enhanced method for the feature extraction of vocal source signals and score combination using an MCE-Based weight estimation of the score of multiple feature vectors are proposed for the performance improvement of speaker recognition systems. The proposed feature vector is composed of perceptual linear predictive cepstral coefficients, skewness, and kurtosis extracted with lowpass filtered glottal flow signals to eliminate the flat spectrum region, which is a meaningless information section. The proposed feature was used to improve the conventional speaker recognition system utilizing the mel-frequency cepstral coefficients and the perceptual linear predictive cepstral coefficients extracted with the speech signals and Gaussian mixture models. In addition, to increase the reliability of the estimated scores, instead of estimating the weight using the probability distribution of the convectional score, the scores evaluated by the conventional vocal tract, and the proposed feature are fused by the MCE-Based score combination method to find the optimal speaker. The experimental results showed that the proposed feature vectors contained valid information to recognize the speaker. In addition, when speaker recognition is performed by combining the MCE-based multiple feature parameter scores, the recognition system outperformed the conventional one, particularly in low Gaussian mixture cases.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.169-171
• /
• 2001

이 논문에서 논해지는 음성변환이라는 것은 어떤 화자의 음성(소스)을 다른 화자의 음성(타겟)으로 바꾸는 것이다. 이 때, 모든 음소들을 녹음해서 데이터베이스화한 음성끼리 매칭시키는 것이 아니라, 몇 번의 학습을 통하여 음색의 특징을 파악한 후 나온 변환함수를 이용하여 원래 화자의 음성을 타겟 음성으로 변환하는 시스템을 제안하고자 한다. 여기서, 음색의 특징들을 추출한 후, 변환함수를 만들기 위한 트레이닝을 위한 방법으로 Gaussian Mixture Modeling을 이용할 것이다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.05a
• /
• pp.682-683
• /
• 2017

주변의 배경잡음으로부터 음성인식률을 향상시키기 위하여 적절한 음성의 특징 파라미터를 선택하는 것이 매우 중요하다. 본 논문에서는 위너필터 방법이 적용된 인간의 청각 특성을 이용한 멜 주파수 켑스트럼 계수를 사용한다. 제안한 멜 주파수 켑스트럼 계수의 특징 파라미터를 다층 퍼셉트론 네트워크에 입력하여 학습시킴으로써 화자인식을 구현한다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2006.11a
• /
• pp.91-94
• /
• 2006

본 논문은 생체인식시스템에서 단일시스템의 각각의 특징을 바탕으로 신뢰성을 증가시키는 것에 있다. 간단하면서 높은 인식률을 가지는 지문과 개개인의 음성을 다중생체인식에 활용하여 다중생체인식 시스템을 구현 하였다. 화자인식부에서는 DSP를 이용하여 화자인식을 수행하고, 이후 지문인식부에서 지문 특징점을 추출하여 KSOM신경망 알고리즘을 이용하여 인식을 수행하였다. 그리고 각 인식부의 전체적인 제어는 ATmega16L을 사용하였다. 또한 인증결과를 PC에 MFC로 디스플레이 한다.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.7
• /
• pp.410-415
• /
• 2005

Recently, most of the speaker verification systems are based on the pattern recognition approach method. And performance of the pattern-classifier depends on how to classify a variety of speakers' feature parameters. In order to classify feature parameters efficiently and effectively, it is of great importance to enlarge variations between speakers and effectively measure distances between feature parameters. Therefore, this paper would suggest the positively mixed model scheme that can enlarge inter-speaker variation by searching the individual model with world model at the same time. During decision procedure, we can maximize inter-speaker variation by using the proposed mixed model scheme. We also make use of a symbol probability weighting function in this system so as to reduce vector quantization errors by measuring symbol probability derived from the distance rate of between the world codebook and individual codebook. As the result of our experiment using this method, we could halve the Detection Cost Function (DCF) of the system from $2.37\%\;to\;1.16\%$.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.151-154
• /
• 2002

본 논문에서는 음성 데이터에서 동일한 화자의 음성 구간을 찾아내는 화자 인덱싱(Speaker Indexing) 기술 중 사전 화자 모델링 과정을 통한 인덱싱 방법을 제안하고 실험하였다. 제안한 인덱싱 방법은 문장 독립(Text Independent) 화자 식별(Speaker Identification)에 사용할 수 있는 모음(Vowel)에 대해 특징 파라미터를 추출하고, 이를 바탕으로 화자별 모델을 구성하였다. 인덱싱은 음성 구간에서 모음의 위치를 검출하고, 구성한 화자 모델과의 거리 계산을 통하여 가장 가까운 모델을 식별된 결과로 한다. 그리고 식별된 결과는 화자 구간 변화와 음성 데이터의 특성을 바탕으로 필터링 과정을 거쳐 최종적인 인덱싱 결과를 얻는다. 화자 인덱싱 실험 대상으로 방송 뉴스를 녹음하여 10명의 화자 모델을 구성하였고, 인덱싱 실험을 수행한 결과 $91.8\%$의 화자 인덱싱 성능을 얻었다.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.497-504
• /
• 2020

Constructing speaker embeddings in speaker verification is an important issue. In general, a self-attention mechanism has been applied for speaker embedding encoding. Previous studies focused on training the self-attention in a high-level layer, such as the last pooling layer. In this case, the effect of low-level layers is not well represented in the speaker embedding encoding. In this study, we propose Masked Cross Self-Attentive Encoding (MCSAE) using ResNet. It focuses on training the features of both high-level and low-level layers. Based on multi-layer aggregation, the output features of each residual layer are used for the MCSAE. In the MCSAE, the interdependence of each input features is trained by cross self-attention module. A random masking regularization module is also applied to prevent overfitting problem. The MCSAE enhances the weight of frames representing the speaker information. Then, the output features are concatenated and encoded in the speaker embedding. Therefore, a more informative speaker embedding is encoded by using the MCSAE. The experimental results showed an equal error rate of 2.63 % using the VoxCeleb1 evaluation dataset. It improved performance compared with the previous self-attentive encoding and state-of-the-art methods.