• ETRI Journal
• /
• 제24권6호
• /
• pp.469-472
• /
• 2002

Speaker adaptation techniques are generally used to reduce speaker differences in speech recognition. In this work, we focus on the features fitted to a linear regression-based speaker adaptation. These are obtained by feature transformation based on independent component analysis (ICA), and the feature transformation matrices are estimated from the training data and adaptation data. Since the adaptation data is not sufficient to reliably estimate the ICA-based feature transformation matrix, it is necessary to adjust the ICA-based feature transformation matrix estimated from a new speaker utterance. To cope with this problem, we propose a smoothing method through a linear interpolation between the speaker-independent (SI) feature transformation matrix and the speaker-dependent (SD) feature transformation matrix. From our experiments, we observed that the proposed method is more effective in the mismatched case. In the mismatched case, the adaptation performance is improved because the smoothed feature transformation matrix makes speaker adaptation using noisy speech more robust.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 1999년도 추계학술대회-지능형 정보기술과 미래조직 Information Technology and Future Organization
• /
• pp.323-335
• /
• 1999

In this paper, we propose a new hybrid model of artificial neural networks(ANNs) and genetic algorithm (GA) to optimal feature transformation and feature weighting. Previous research proposed several variants of hybrid ANNs and GA models including feature weighting, feature subset selection and network structure optimization. Among the vast majority of these studies, however, ANNs did not learn the patterns of data well, because they employed GA for simple use. In this study, we incorporate GA in a simultaneous manner to improve the learning and generalization ability of ANNs. In this study, GA plays role to optimize feature weighting and feature transformation simultaneously. Globally optimized feature weighting overcome the well-known limitations of gradient descent algorithm and globally optimized feature transformation also reduce the dimensionality of the feature space and eliminate irrelevant factors in modeling ANNs. By this procedure, we can improve the performance and enhance the generalisability of ANNs.

• ETRI Journal
• /
• 제34권6호
• /
• pp.847-857
• /
• 2012

Intrusion detection systems (IDSs) have an important effect on system defense and security. Recently, most IDS methods have used transformed features, selected features, or original features. Both feature transformation and feature selection have their advantages. Neighborhood component analysis feature transformation and genetic feature selection (NCAGAFS) is proposed in this research. NCAGAFS is based on soft computing and data mining and uses the advantages of both transformation and selection. This method transforms features via neighborhood component analysis and chooses the best features with a classifier based on a genetic feature selection method. This novel approach is verified using the KDD Cup99 dataset, demonstrating higher performances than other well-known methods under various classifiers have demonstrated.

• 한국컴퓨터정보학회논문지
• /
• 제13권5호
• /
• pp.53-67
• /
• 2008

현재 유사 도메인에 대한 feature 모델과 class 모델간의 재사용을 위해, 모델 차원에서 상호변환 연구와 두 모델간 온톨로지를 이용한 변환 연구가 있으나, 메타모델을 통한 일관성 있는 변환이 되지 못하며, 각 모델이 가진 변환 대상 모델링 요소가 충분치 않고, 특히, 자동 변환 알고리즘 및 지원 툴을 제공하지 않음으로써 모델간 재사용의 저하를 초래하고 있다. 본 논문에서는 메타모델 상에서 온톨로지를 사용한 feature 모델을 class 모델로의 변환 방법을 제시한다. 이를 위해, feature 모델, class 모델 및 온톨로지에 대한 메타모델을 재정의하고, 각 메타모델별 모델링 요소에 대한 속성을 정의한다. 이 속성들에 기반하여 feature 모델과 온톨로지 간 그리고 온톨로지와 class 모델간의 변환 규칙 프로파일을 집합 이론과 명제논리로 정의한다. 이러한 변환의 자동화 구축을 위해 변환 알고리즘을 생성하고, 지원 툴을 구현한다. 제시한 변환규칙 및 툴을 사용해 전자 결재시스템을 통해 실제 적용한다. 기대효과로써, 기 구축된 feature 모델을 class모델로 변환하여 상이한 개발방법간에 생성된 모델을 재사용을 할 수 있다. 특히, 온톨로지를 사용해서 의미적 변환의 모호성을 해소시킬 수 있으며, 변환의 자동화 및 모델간 일관성을 유지시켜줄 수 있다.

• 대한음성학회지:말소리
• /
• 제43호
• /
• pp.127-136
• /
• 2002

The speaker adaptation technique is generally used to reduce the speaker difference in speech recognition. In this work, we focus on the features fitted to a linear regression-based speaker adaptation. These are obtained by feature transformation based on independent component analysis (ICA), and the transformation matrix is learned from a speaker independent training data. When the amount of data is small, however, it is necessary to adjust the ICA-based transformation matrix estimated from a new speaker utterance. To cope with this problem, we propose a smoothing method: through a linear interpolation between the speaker-independent (SI) feature transformation matrix and the speaker-dependent (SD) feature transformation matrix. We observed that the proposed technique is effective to adaptation performance.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제8권3호
• /
• pp.192-195
• /
• 2008

People have tendency of forgetting music title, though they easily remember particular part of music. If a music search system can find the title through a part of melody, this will provide very convenient interface to users. In this paper, we propose an algorithm that enables this type of search using feature transformation function. The original music is transformed to new feature information with sequential melodies. When a melody that is a part of search music is given to the system, the music retrieval system searches the music similar to the feature information of the melody. Moreover, this transformation function can be easily extended to various music recognition systems.

• The Journal of the Acoustical Society of Korea
• /
• 제22권4E호
• /
• pp.150-154
• /
• 2003

The distortion effects of microphones have been analyzed and compensated at mel-frequency feature domain. Unlike popular bias removal algorithms a linear transformation of mel-frequency spectrum is incorporated. Although a diagonal matrix transformation is sufficient for medium-quality microphones, a full-matrix transform is required for low-quality microphones with severe nonlinearity. Proposed compensation algorithms are tested with HTIMIT database, which resulted in about 5 percents improvements in recognition rate over conventional CMS algorithm.

• 한국정밀공학회지
• /
• 제10권4호
• /
• pp.30-41
• /
• 1993

Developed in this paper is a feature based modeling system for the design of welded plat construction(WPC) which is composed of flat or bended plates represented as reference plane with a constant thickness. First, the necessity and the characteristics of the modeing system for WPC as compared with the assembly of mechanical parts are investigated. Secondly, feature library for the assembly of WPC is shown which contains several types of features like joint feature, groove feature, material feature, and precision feature. Thirdly, the assembly procedures are presented which mainly consist of both the assembly transformation and the correct assembly checking. Fourthly, weld lines of the assembled WPC are defined so that those can be used in the process planning or the manufacturing stage. Finally, a prototype by a geometric modeling software Pro/Engineer, a graphic software GL(Graphic Library), and C language on a CAD workstation IRIS.

• 말소리와 음성과학
• /
• 제1권4호
• /
• pp.127-132
• /
• 2009

The feature vectors which are used in conventional speaker recognition (SR) systems may have many correlations between their neighbors. To improve the performance of the SR, many researchers adopted linear transformation method like principal component analysis (PCA). In general, the linear transformation of the feature vectors is based on concatenated form of the static features and their dynamic features. However, the linear transformation which based on both the static features and their dynamic features is more complex than that based on the static features alone due to the high order of the features. To overcome these problems, we propose an efficient method that applies linear transformation and temporal information of the features to reduce complexity and improve the performance in speaker verification (SV). The proposed method first performs a linear transformation by PCA coefficients. The delta parameters for temporal information are then obtained from the transformed features. The proposed method only requires 1/4 in the size of the covariance matrix compared with adding the static and their dynamic features for PCA coefficients. Also, the delta parameters are extracted from the linearly transformed features after the reduction of dimension in the static features. Compared with the PCA and conventional methods in terms of equal error rate (EER) in SV, the proposed method shows better performance while requiring less storage space and complexity.

• 한국CDE학회논문집
• /
• 제12권5호
• /
• pp.354-365
• /
• 2007

We present novel methods to generate a sequence of shapes that represents the pattern of morphological development or transformation of Bezier curves. The presented methods utilize the intrinsic geometric structures of a Bezier curve that are derived from rib and fan decomposition (RFD). Morphological development based on RFD shows a characteristic pattern of structural growth of a Bezier curve, which is the direct consequence of development path defined by fans. Morphological transformation based RFD utilizes development patterns of source and target curves to mimic the theory of evolutionary developmental biology: although the source and target curves are quite different in shapes, we can easily find similarities in their younger shapes, which makes it easier to set up feature correspondences for blending them. We also show that further controls on base transformation for intensity of feature blending, and extrapolation can compensate the immaturity of blended curves. We demonstrate the experimental results where transformation patterns are smoother and have unique geometric style that cannot be generated using conventional methods based on multi-linear blending.