• The Journal of the Acoustical Society of Korea
• /
• v.34 no.2
• /
• pp.171-176
• /
• 2015

We present a speaker verification method by extracting i-vectors based on dominant state information of Hidden Markov Model (HMM) - Universal Background Model (UBM). Ergodic HMM is used for estimating UBM so that various characteristic of individual speaker can be effectively classified. Unlike Gaussian Mixture Model(GMM)-UBM based speaker verification system, the proposed system obtains i-vectors corresponding to each HMM state. Among them, the i-vector for feature is selected by extracting it from the specific state containing dominant state information. Relevant experiments are conducted for validating the proposed system performance using the National Institute of Standards and Technology (NIST) 2008 Speaker Recognition Evaluation (SRE) database. As a result, 12 % improvement is attained in terms of equal error rate.

• Phonetics and Speech Sciences
• /
• v.7 no.1
• /
• pp.3-10
• /
• 2015

In this paper, different frequency scales in cepstral feature extraction are evaluated for the text-independent speaker recognition. To this end, mel-frequency cepstral coefficients (MFCCs), linear frequency cepstral coefficients (LFCCs), and bilinear warped frequency cepstral coefficients (BWFCCs) are applied to the speaker recognition experiment. In addition, the spectro-temporal features extracted by the cepstral-time matrix (CTM) are examined as an alternative to the delta and delta-delta features. Experiments on the NIST speaker recognition evaluation (SRE) 2004 task are carried out using the Gaussian mixture model-universal background model (GMM-UBM) method and the joint factor analysis (JFA) method, both based on the ALIZE 3.0 toolkit. Experimental results using both the methods show that BWFCC with appropriate warping factor yields better performance than MFCC and LFCC. It is also shown that the feature set including the spectro-temporal information based on the CTM outperforms the conventional feature set including the delta and delta-delta features.

• Proceedings of the Korea Contents Association Conference
• /
• 2016.05a
• /
• pp.11-12
• /
• 2016

SVM (Support Vector machine) is powerful machine-learning method, and obtains better performance than traditional methods in the applications of muti-dimension nonlinear pattern classification. For the case of SVM model training and low efficiency in large samples, this paper proposes a combination of statistical parameters of the GMM-UBM (Universal Background Model) model. It is very effective to solve the problem of the large sample for the SVM training. The experiment is carried on four special dynamic hand gestures using the EPIC sensors. And the results show that the improved dynamic hand gesture recognition system has a high recognition rate up to 96.75%.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.06a
• /
• pp.155-156
• /
• 2013

소셜 TV 사용 시, 사용자들은 TV를 시청하면서 타 사용자와의 소통을 위해 리모컨을 이용해서 텍스트를 작성해야하는 불편함을 가지고 있다. 본 논문에서는 소셜 TV의 이러한 불편함을 해결하기 위해 사용자 반응 사운드를 자동으로 인식하여 상대방에게 이모티콘을 전달하기 위한 시스템을 제안하며, 사용자 반응 사운드 인식에 사용되는 분류방식들을 비교한다. 사용자 반응 사운드 인식을 위해 사용되는 분류 방식들 중에서, Gaussian Mixture Model(GMM), Gaussian Mixture Model - Universal Background Model(GMM-UBM), Hidden Markov Model(HMM), Support Vector Machine(SVM)의 성능을 비교하였다. 각 분류기의 성능을 비교하기 위하여 MFCC 특징값을 각 분류기에 적용하여 사용자 반응 사운드 인식에 가장 최적화된 분류기를 선택하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.1
• /
• pp.36-42
• /
• 2006

A Principal component analysis (PCA)-based Mel-Frequency Discrete Wavelet Coefficients (MFDWC) feature Parameters for speaker verification system is Presented in this Paper In this method, we used the 1st-eigenvector obtained from PCA to calculate the energy of each node of level that was approximated by. met-scale. This eigenvector satisfies the constraint of general weighting function that the squared sum of each component of weighting function is unity and is considered to represent speaker's characteristic closely because the 1st-eigenvector of each speaker is fairly different from the others. For verification. we used Universal Background Model (UBM) approach that compares claimed speaker s model with UBM on frame-level. We performed experiments to test the effectiveness of PCA-based parameter and found that our Proposed Parameters could obtain improved average Performance of $0.80\%$compared to MFCC. $5.14\%$ to LPCC and 6.69 to existing MFDWC.

• Phonetics and Speech Sciences
• /
• v.6 no.2
• /
• pp.21-28
• /
• 2014

We propose a novel scheme for digital audio authentication of given audio files which are edited by inserting small audio segments from different environmental sources. The purpose of this research is to detect inserted sections from given audio files. We expect that the proposed method will assist human investigators by notifying suspected audio section which considered to be recorded or transmitted on different environments. GMM-UBM and GSV-SVM are applied for modeling the dominant environment of a given audio file. Four kinds of likelihood ratio based scores and SVM score are used to measure the likelihood for a dominant environment model. We also use an ensemble score which is a combination of the aforementioned five kinds of scores. In the experimental results, the proposed method shows the lowest average equal error rate when we use the ensemble score. Even when dominant environments were unknown, the proposed method gives a similar accuracy.

• ETRI Journal
• /
• v.28 no.6
• /
• pp.811-814
• /
• 2006

In this letter, we introduce confusion-based confidence measures for detecting an impostor in speaker recognition, which does not require an alternative hypothesis. Most traditional speaker verification methods are based on a hypothesis test, and their performance depends on the robustness of an alternative hypothesis. Compared with the conventional Gaussian mixture model-universal background model (GMM-UBM) scheme, our confusion-based measures show better performance in noise-corrupted speech. The additional computational requirements for our methods are negligible when used to detect or reject impostors.

• Journal of Information Processing Systems
• /
• v.13 no.4
• /
• pp.928-940
• /
• 2017

State-of-the-art speaker recognition systems may work better for the English language. However, if the same system is used for recognizing those who speak different languages, the systems may yield a poor performance. In this work, the decisions of a Gaussian mixture model-universal background model (GMM-UBM) and a learning vector quantization (LVQ) are combined to improve the recognition performance of a multilingual speaker identification system. The difference between these classifiers is in their modeling techniques. The former one is based on probabilistic approach and the latter one is based on the fine-tuning of neurons. Since the approaches are different, each modeling technique identifies different sets of speakers for the same database set. Therefore, the decisions of the classifiers may be used to improve the performance. In this study, multitaper mel-frequency cepstral coefficients (MFCCs) are used as the features and the monolingual and cross-lingual speaker identification studies are conducted using NIST-2003 and our own database. The experimental results show that the combined system improves the performance by nearly 10% compared with that of the individual classifier.

• Journal of Convergence for Information Technology
• /
• v.7 no.6
• /
• pp.159-166
• /
• 2017

In this paper, we have experimented to improve the verification performance of speaker change detection on broadcast news. It is to enhance the input noisy speech and to apply the KL distance $D_s$ using the SNR-based weighting function $w_m$. The basic experimental system is the verification system of speaker change using GMM-UBM based KL distance D(Experiment 0). Experiment 1 applies the input noisy speech enhancement using MMSE Log-STSA. Experiment 2 applies the new KL distance $D_s$ to the system of Experiment 1. Experiments were conducted under the condition of 0% MDR in order to prevent missing information of speaker change. The FAR of Experiment 0 was 71.5%. The FAR of Experiment 1 was 67.3%, which was 4.2% higher than that of Experiment 0. The FAR of experiment 2 was 60.7%, which was 10.8% higher than that of experiment 0.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.3
• /
• pp.279-284
• /
• 2009

Speaker verification determines whether the claimed speaker is accepted based on the score of the test utterance. In recent years, methods based on Gaussian mixture models and universal background model have been the dominant approaches for text-independent speaker verification. These speaker verification systems based on these methods provide very good performance under laboratory conditions. However, in real situations, the performance of speaker verification system is degraded dramatically. For overcoming this performance degradation, the feature recombination method was proposed, but this method had a drawback that whole sub-band feature vectors are used to compute the likelihood scores. To deal with this drawback, a modified feature recombination method which can use each sub-band likelihood score independently was proposed in our previous research. In this paper, we propose a sub-band weighting method based on sub-band signal-to-noise ratio which is combined with previously proposed modified feature recombination. This proposed method reduces errors by 28% compared with the conventional feature recombination method.