• Proceedings of the Acoustical Society of Korea Conference
• /
• 한국음향학회 2000년도 하계학술발표대회 논문집 제19권 1호
• /
• pp.55-58
• /
• 2000

본 논문에서는 화자 확인 시스템의 등록과 확인 과정의 채널 환경 불일치로 성능이 저하되는 문제를 해결하기 위한 새로운 정규화 방법에 대해 설명한다. 제안된 방법은 첫째, 입력 음성으로부터 효과적으로 채널을 추정$\cdot$보상하고 둘째, 스코어 정규화 과정에서 사칭자 모델로서 사용되는 world모델과의 차이를 채널 추정 및 화자 모델 생성에 효과적으로 사용하는 것을 목표로 한다. 이를 위해 입력 음성의 켑스트럼과 HMM world 모델의 파라메터인 평균 켑스트럼과의 차이를 통해 음소열에 종속적인 채널 켑스트럼인 Phone-Dependent Difference Cepstrum을 추정한다. 한편 입력 음성의 음소열은 world모델의 스코어를 얻는 과정에서 함께 얻어질 수 있다. 채널 추정 실험 결과를 통해서 가장 일반적인 채널 정규화방법인 CMS에 의해 추정된 채널에 비해 실제 채널과 유사하며 화자 고유의 특성을 왜곡시키지 않는 채널 추정이 가능함을 확인할 수 있었다.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• 제14권3호
• /
• pp.208-214
• /
• 2021

Recently, much of the intelligent security scenario and criminal investigation demands for matching photo and non-photo. Existing face recognition system can not sufficiently guarantee these needs. In this paper, we propose an algorithm to improve the performance of heterogeneous face recognition systems by reducing the different modality between sketches and photos of the same person. The proposed algorithm extracts each image's texture features through texture descriptors (gray level co-occurrence matrix, multiscale local binary pattern), and based on this, generates a transformation matrix through eigenfeature regularization and extraction techniques. The score value calculated between the vectors generated in this way finally recognizes the identity of the sketch image through the score normalization methods.

• Proceedings of the IEEK Conference
• /
• 대한전자공학회 2001년도 제14회 신호처리 합동 학술대회 논문집
• /
• pp.487-490
• /
• 2001

본 논문에서는 기존의 Caussian Mixture Model을 이용한 실시간문맥독립화자인식시스템의 성능을 향상시키기 위하여 화자검증시스템에서 좋은 결과를 나타내는 유사도정규화 ( Likelihood Normalization )방법을 화자식별시스템에 적용하여 시스템을 구현하였으며, 인식실험한 결과에 대해 보고한다. 시스템은 화자모델생성단과 화자식별단으로 구성하였으며, 화자모델생성단에서는, 화자발성의 음향학적 특징을 잘 표현할 수 있는 GMM(Gaussian Mixture Model)을 이용하여 화자모델을 작성하였으며. GMM의 파라미터를 최적화하기 위하여 MLE(Maximum Likelihood Estimation)방법을 사용하였다. 화자식별단에서는 학습된 데이터와 테스트용 데이터로부터 ML(Maximum Likelihood)을 이용하여 프레임단위로 유사도를 계산하였다. 계산된 유사도는 유사도 정규화 과정을 거쳐 스코어( SC)로 표현하였으며, 가장 높은 스코어를 가지는 화자를 인식화자로 결정한다. 화자인식에서 발성의 종류로는 문맥독립 문장을 사용하였다. 인식실험을 위해서는 ETRI445 DB와 KLE452 DB를 사용하였으며. 특징파라미터로서는 켑스트럼계수 및 회귀계수값만을 사용하였다. 인식실험에서는 등록화자의 수를 달리하여 일반적인 화자식별방법과 프레임단위유사도정규화방법으로 각각 인식실험을 하였다. 인식실험결과, 프레임단위유사도정규화방법이 인식화자수가 많아지는 경우에 일반적인 방법보다 향상된 인식률을 얻을수 있었다.

• Journal of the Institute of Convergence Signal Processing
• /
• 제3권1호
• /
• pp.8-14
• /
• 2002

In this paper, we realized a real-time text-independent speaker recognition system using gaussian mixture model, and applied frame level likelihood normalization method which shows its effects in verification system. The system has three parts as front-end, training, recognition. In front-end part, cepstral mean normalization and silence removal method were applied to consider speaker's speaking variations. In training, gaussian mixture model was used for speaker's acoustic feature modeling, and maximum likelihood estimation was used for GMM parameter optimization. In recognition, likelihood score was calculated with speaker models and test data at frame level. As test sentences, we used text-independent sentences. ETRI 445 and KLE 452 database were used for training and test, and cepstrum coefficient and regressive coefficient were used as feature parameters. The experiment results show that the frame-level likelihood method's recognition result is higher than conventional method's, independently the number of registered speakers.

• The Journal of the Acoustical Society of Korea
• /
• 제22권5호
• /
• pp.383-387
• /
• 2003

This paper deals with the method of speaker verification based on the conventional cohort of fixed size. In particular, a new cohort of variable size, which makes use of the distance between speaker models, is proposed: The density of neighboring speaker models within the fixed distance from each speaker is taken into account in the proposed method. The high density leads to the increase of cohort size, thus improving the speaker verification rate. On the other hand, the low density leads to its decrease, thus reducing the amount of computations. The simulation results show that the proposed method outperforms the conventional one, achieving a reduction in the EER.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• 제45권2호
• /
• pp.1-9
• /
• 2008

As the security requirements of mobile phones have been increasing, there have been extensive researches using one biometric feature (e.g., an iris, a fingerprint, or a face image) for authentication. Due to the limitation of uni-modal biometrics, we propose a method that combines face and iris images in order to improve accuracy in mobile environments. This paper presents four advantages and contributions over previous research. First, in order to capture both face and iris image at fast speed and simultaneously, we use a built-in conventional mega pixel camera in mobile phone, which is revised to capture the NIR (Near-InfraRed) face and iris image. Second, in order to increase the authentication accuracy of face and iris, we propose a score level fusion method based on SVM (Support Vector Machine). Third, to reduce the classification complexities of SVM and intra-variation of face and iris data, we normalize the input face and iris data, respectively. For face, a NIR illuminator and NIR passing filter on camera are used to reduce the illumination variance caused by environmental visible lighting and the consequent saturated region in face by the NIR illuminator is normalized by low processing logarithmic algorithm considering mobile phone. For iris, image transform into polar coordinate and iris code shifting are used for obtaining robust identification accuracy irrespective of image capturing condition. Fourth, to increase the processing speed on mobile phone, we use integer based face and iris authentication algorithms. Experimental results were tested with face and iris images by mega-pixel camera of mobile phone. It showed that the authentication accuracy using SVM was better than those of uni-modal (face or iris), SUM, MAX, NIN and weighted SUM rules.

• Journal of KIISE:Software and Applications
• /
• 제37권1호
• /
• pp.1-8
• /
• 2010

Recently lots of attention has been paid to gene set analysis for identifying differentially expressed gene-sets between two sample groups. Unlike earlier approaches, the gene set analysis enables us to find significant gene-sets along with their functional characteristics. For this reason, various novel approaches have been suggested lately for gene set analysis. As one of such, PAGE is a parametric approach that employs average difference (AD) as an expression metric to quantify expression differences between two sample groups and assumes that the distribution of gene scores is normal. This approach is preferred to non-parametric approach because of more effective performance. However, the metric AD does not reflect either gene expression intensities or variances over samples in calculating gene scores. Thus, in this paper, we investigate the usefulness of several other expression metrics for parametric gene-set analysis, which consider actual expression intensities of genes or their expression variances over samples. For this purpose, we examined three expression metrics, WAD (weighted average difference), FC (Fisher's criterion), and Abs_SNR (Absolute value of signal-to-noise ratio) for parametric gene set analysis and evaluated their experimental results.

• Horticultural Science & Technology
• /
• 제28권5호
• /
• pp.836-844
• /
• 2010

This study was carried out to investigate plant characteristics and nutritional components of the genetically modified (GM) Chinese cabbage and its control line grown in the central and northern parts of Korea in order to establish the evaluating protocol and standard assessment. The GM and non-GM Chinese cabbage was planted with normal and concentrated density at two locations in spring and fall of 2008 and 2009. From the statistic analysis on plant characteristics and nutritional components, there were not many significant differences between GM and non-GM Chinese cabbage. Only few differences in the plant characteristics were found between the dense and normal planting. In the dense planting, there was no significant difference between GM and non-GM Chinese cabbages except for three out of 18 plant traits, such as leaf shape, hairiness and midrib length. On the other hand, nine plant traits including leaf length, leaf width, leaf color, leaf shape, fresh weigh of ground part, number of leaf, midrib length, midrib width and root diameter were slightly different between GM and non-GM Chinese cabbage in the normal planting. In case of leaf length, midrib length, midrib width and fresh weigh of ground part, there were significantly differences not only between two lines, but also between two locations. From nutritional component analysis, only five fatty acids were identified in the Chinese cabbage: palmitic acid, oleic acid, stearic acid, linoleic acid and linolenic acid. Except linoleic acid, four fatty acids in one gram of dried sample from GM line were little higher than those from non-GM line. However, there were no significant differences in total contents of fatty acids not only between GM and non-GM Chinese cabbage line, but also between northern and central cultivating areas in the normal and dense planting. According to the composition of inorganic elements identified in the samples from both lines, there were six macro-elements, such as N, P, Ca, K, Mg and Na, and four micro-elements, Cu, Fe, Mn and Zn. Based on the result from PCA analysis, specific clusters were not found between GM Chinese cabbage and the control line, but found between two regions.