• The Journal of the Acoustical Society of Korea
• /
• v.38 no.5
• /
• pp.593-600
• /
• 2019

The similarity in tones between speakers can lower the performance of speaker verification. To improve the performance of speaker verification systems, we propose a multi-task learning technique using deep neural network to learn speaker information and age information. Multi-task learning can improve generalization performances, because it helps deep neural networks to prevent hidden layers from overfitting into one task. However, we found in experiments that learning of age information does not work well in the process of learning the deep neural network. In order to improve the learning, we propose a method to dynamically change the objective function weights of speaker identification and age estimation in the learning process. Results show the equal error rate based on RSR2015 evaluation data set, 6.91 % for the speaker verification system without using age information, 6.77 % using age information only, and 4.73 % using age information when weight change technique was applied.

• 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.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.5
• /
• pp.69-72
• /
• 1998

본 논문에서는 문장독립형 화자인식 시스템에 MQ(Matrix Quantization) 방법 사용 을 제안한다. 또한 인식율을 높이기 위해 MQ를 수정한 방법인 SMQ(Separated Matrix Quantization)을 제안한다. 기존의 VQ-distortion 방법은 대체로 좋은 성능을 가지나 화자의 동적 특성을 이용하지 못한다는 단점이 있다. MQ와 SMQ는 화자의 동적 특성을 이용할 수 있으므로 시간 변화에 대한 화자의 특징 변화까지 모델링 할 수 있는 장점이 있다. MQ는 여러 프레임을 묶어 Matrix Codebook을 가지며 SMQ는 MQ의 기본 codebook을 다시 켑스 트럼의 차수에 따라 나누어 codebook을 만든다. 즉, 켑스트럼 차수를 저, 중, 고차로 나누어 각 부분별로 Matrix codebook을 만들도록 한다. 인식실험은 문장독립 음성 데이터에 대해 실행했으며 MQ모델의 경우 Matrix의 크기를 짧은 음소크기부터 음절단위까지 변화시켜 실 험하였다. 아울러 SMQ 모델에서의 실험은 차수별 유용도를 보기 위하여 부분 차수를 이용 하여 실험하였다. 실험결과 MQ와 SMQ방법이 VQ에 비해 좋은 성능을 가짐을 확인하였다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06c
• /
• pp.131-134
• /
• 1998

본 논문에서는 전자 개인 비서 시스템(PEAS)의 일부인 일정관리 시스템을 화자인식 기술을 적용하여 구현하였다. 본 시스템은 음성을 패스워드로 개인을 확인하여 각 개인의 일정을 관리해 주는 것으로 보안성과 함께 사용자에게 편의성을 제공한다. 사용자 등록을 자유롭게 하였으며 인식에서는 계산 시간 등을 고려하여 DTW 알고리즘에서 얻을 수 있는 경로정보를 이용해 하나의 참조패턴을 구성하도록 하였다. 또한 시간 흐름에 따라 인식율 저하를 방지하기 위해 실험결과에 따라 일정기간 뒤에 자동으로 참조패턴이 갱신되도록 하였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.3
• /
• pp.348-354
• /
• 2003

This proposes a new algorithm for advanced personal identification system using iris pattern and speech signal. Since the proposed algorithm adopts a fusion scheme to take advantage of iris recognition and speaker identification, it shows robustness for noisy environments. For evaluating the performance of the proposed scheme, we compare it with the iris pattern recognition and speaker identification respectively. In the experiments, the proposed method showed more 56.7% improvements than the iris recognition method and more 10% improvements than the speaker identification method for high quality security level. Also, in noisy environments, the proposed method showed more 30% improvements than the iris recognition method and more 60% improvements than the speaker identification method for high quality security level.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.6
• /
• pp.26-33
• /
• 2001

Though speaker recognition based on conventional AR HMM shows good performance, its lack of modeling the environmental noise makes its performance degraded in case of practical noisy environment. In this paper, a robust speaker recognition system based on AR HMM is proposed, where noise is considered in the observation signal model for practical noisy environment and duration-term is considered to increase performance. Experimental results, using the digits database from 100 speakers (77 males and 23 females) under white noise and car noise, show improved performance.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.4
• /
• pp.22-28
• /
• 2001

This paper describes the implementation of a robust speaker position location system using the voice signal received by microphone array. To be robust to the reverberation which is the major factor of the performance degradation, low-frequency phase restoration algorithm which eliminates the influence of reverberations using the low-frequency information of the CPSP function is proposed. The implemented real-time system consists of a general purpose DSP (TMS320C31 of Texas instruments), analog part which contains amplifiers and filters, and digital part which is composed of the external memory and 12-bit A/D converter. In the real conference room environment, the implemented system that was constructed by the proposed algorithms showed better performance than the conventional system. The error of the TDOA estimation reduced more than 15 samples.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.4
• /
• pp.299-307
• /
• 2003

There have been many studies on speaker normalization which aims to minimize the effects of speaker's vocal tract length on the recognition performance of the speaker independent speech recognition system. In this paper, we propose a simple vector quantizer based linear warping speaker normalization method based on the observation that the vector quantizer can be successfully used for speaker verification. For this purpose, we firstly generate an optimal codebook which will be used as the basis of the speaker normalization, and then the warping factor of the unknown speaker will be extracted by comparing the feature vectors and the codebook. Finally, the extracted warping factor is used to linearly warp the Mel scale filter bank adopted in the course of MFCC calculation. To test the performance of the proposed method, a series of recognition experiments are conducted on discrete HMM with thirteen mono-syllabic Korean number utterances. The results showed that about 29% of word error rate can be reduced, and that the proposed warping factor extraction method is useful due to its simplicity compared to other line search warping methods.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.1015-1020
• /
• 2003

Among the techniques to protect private information by adopting biometrics, speaker verification is expected to be widely used due to advantages in convenient usage and implementation cost. Speaker verification should achieve a high degree of the reliability in the verification score, the flexibility in speech text usage, and the efficiency in verification system complexity. Continuants have excellent speaker-discriminant power and the modest number of phonemes in the category, and multilayer perceptrons (MLPs) have superior recognition ability and fast operation speed. In consequence, the two provide viable ways for speaker verification system to obtain the above properties. This paper implements a system to which continuants and MLPs are applied, and evaluates the system using a Korean speech database. The results of the experiment prove that continuants and MLPs enable the system to acquire the three properties.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.3
• /
• pp.228-234
• /
• 2004

In this paper, we proposed the compensation technique that overcomes the limitations of the conventional approaches through summing up the bias terms between world's codebook and individual codebook vectors of feature parameters. But, mean compensation without condition can bring higher false acceptance. Therefore, the proposed technique compensates the channel mis-match condition by weighted bias sum using nonlinear function regarding to the distortion between speech and silence. The simulation results show that the FRR (flase reject rate) is decreased 14.95% when the proposed algorithm was applied.