• Journal of the Korean Data and Information Science Society
• /
• v.7 no.1
• /
• pp.113-118
• /
• 1996

Principal component analysis(PCA) is an essential technique for data compression and feature extraction, and has been widely used in statistical data analysis, communication theory, pattern recognition, and image processing. Oja(1992) found that a linear neuron with constrained Hebbian learning rule can extract the principal component by using stochastic gradient ascent method. In practice real data often contain some outliers. These outliers will significantly deteriorate the performances of the PCA algorithms. In order to make PCA robust, Xu & Yuille(1995) applied statistical physics to the problem of robust principal component analysis(RPCA). Devlin et.al(1981) obtained principal components by using techniques such as M-estimation. The propose of this paper is to investigate from the statistical point of view how Xu & Yuille's(1995) RPCA works under the same simulation condition as in Devlin et.al(1981).

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.7
• /
• pp.519-527
• /
• 2003

Speech is much influenced by the existence of outliers which are introduced by such an unexpected happenings as additive background noise, change of speaker's utterance pattern and voice detection errors. These kinds of outliers may result in severe degradation of speaker recognition performance. In this paper, we proposed the GMM based on robust principal component analysis (RPCA-GMM) using M-estimation to solve the problems of both ouliers and high dimensionality of training feature vectors in speaker identification. Firstly, a new feature vector with reduced dimension is obtained by robust PCA obtained from M-estimation. The robust PCA transforms the original dimensional feature vector onto the reduced dimensional linear subspace that is spanned by the leading eigenvectors of the covariance matrix of feature vector. Secondly, the GMM with diagonal covariance matrix is obtained from these transformed feature vectors. We peformed speaker identification experiments to show the effectiveness of the proposed method. We compared the proposed method (RPCA-GMM) with transformed feature vectors to the PCA and the conventional GMM with diagonal matrix. Whenever the portion of outliers increases by every 2%, the proposed method maintains almost same speaker identification rate with 0.03% of little degradation, while the conventional GMM and the PCA shows much degradation of that by 0.65% and 0.55%, respectively This means that our method is more robust to the existence of outlier.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.11a
• /
• pp.261-264
• /
• 2003

얼굴인식은 동일 사람의 얼굴이라도 조명변화나 얼굴 표정변화에 따라 매우 다른 영상들로 나타나기 때문에 매우 어려운 문제이다. 본 논문에서는 조명변화에도 강인하고 얼굴영상에 대해 높은 얼굴 인식률을 얻기 위해 2D-HMM(Hidden Markov Model) 얼굴인식 방법을 제안하고 실험하였다. 제안된 방법은 조명변화에 대해서 조명변화 함수인 $\delta$(delta) 함수를 0, 40, 60, 80으로 변화해 가면서 이미지 보정을 실험하였으며, 계산의 복잡성을 줄이고 얼굴영상에 대한 높은 인식률을 얻기 위해 기존의 픽셀값 대신에 2D-DCT 계수를 관측벡터로 사용하였다. 시스템의 성능을 평가하기 위해 정량적 평가방법은 FAR(False Accpt Rate)와 FRR(False Reject Rate)를 측정하여 비교하였으며, 기존의 얼굴인식 방법인 PCA, 1차원 HMM과 비교분석하였다. 실험결과 2D-HMM의 경우 FAR(False Accept Rate)가 5.08%로 ID-HMM 5.18%, PCA 10.16%보다 높은 성능을 보였으며, FRR(False Reject Rate)의 경우에도 0.01%로 10.16%인 PCA보다 좋은 성능을 보였다. 이로서 조명변화에 대해서는 PCA보다 2D-HMM 얼굴인식 방법이 우수함을 알 수 있었다.

• Proceedings of the IEEK Conference
• /
• 2006.06a
• /
• pp.283-284
• /
• 2006

In this paper, we propose a novel segmentation and classification method using texture features for the document image. First, we extract the local entropy and then segment the document image to separate the background and the foreground using the Otsu's method. Finally, we classify the segmented regions into each component using PCA(principle component analysis) algorithm based on the texture features that are extracted from the co-occurrence matrix for the entropy image. The entropy-based segmentation is robust to not only noise and the change of light, but also skew and rotation. Texture features are not restricted from any form of the document image and have a superior discrimination for each component. In addition, PCA algorithm used for the classifier can classify the components more robustly than neural network.

• Speech Sciences
• /
• v.14 no.2
• /
• pp.127-135
• /
• 2007

Feature vectors of speech are generally extracted from whole frequency domain. The inherent character of a speaker is located in the low band or high band frequency. However, if the speech is corrupted by narrowband noise with concentrated energy, speaker verification performance is reduced as the individual characteristic is removed. In this paper, we propose a PCA Covariance Model based on the multiband to extract the robust feature vectors against the narrowband noise. First, we divide the overall frequency band into several subbands. Second, the correlation of feature vectors extracted independently from each subband is removed by PCA. The distance obtained from each subband has different distribution. To normalize against the different distribution, we moved the value into the normalized distribution through the mapping function. Finally, the represented value applying the weighting function is used for speaker verification. In the experiments, the proposed method shows better performance of the speaker verification and reduces the computation.

• The KIPS Transactions:PartB
• /
• v.11B no.6
• /
• pp.645-652
• /
• 2004

Among the biometrics, signature shows more larger variation than the other biometrics such as fingerprint and iris. In order to overcome this problem, we propose a robust offline signature verification method based on PCA and LDA. Signature is projected to vertical and horizontal axes by new grid partition method. And then feature extraction and decision is performed by PCA and LDA. Experimental results show that the proposed offline signature verification has lower False Reject Rate(FRR) and False Acceptance Rate(FAR) which are 1.45% and 2.1%, respectively.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.225-228
• /
• 2002

본 논문에서는 화자인식을 위하여 강인한 주성분 분석법(Robust Principal Component Analysis)을 갖는 화자인식 방법을 제안하였다. 강인한 주성분 분석법은 특징벡터들의 outlier가 존재할 경우 k-차원으로 줄이면서 강인한 화자 모델을 만들기 위하여 사용한다. 기존의 PCA 방법은 순수한 화자의 정보가 잡음 등의 outlier에 의해 손상될 수 있으므로, 강인한 주성분 분석법을 사용하여 outlier의 영향을 감소 시켰다. 화자 별로 k-차원 diagonal GMM 학습시 mixture 수를 적응시켜 데이터 저장 공간을 최소화하였다. 200명의 고립 숫자음을 사용하여 기존의 diagonal GMM 방법과 제안된 방법을 실험한 결과, 제안된 방법에서 약 $1.5\%$더 높은 인증률을 얻을 수 있었다.

• Journal of Korea Multimedia Society
• /
• v.10 no.1
• /
• pp.26-35
• /
• 2007

We present a robust face tracking system from the sequence of video images based on effective detector and Kalman filter. To construct the effective face detector, we extract the face features using the five types of simple Haar-like features. Extracted features are reinterpreted using Principal Component Analysis (PCA), and interpreted principal components are used for Support Vector Machine (SVM) that classifies the faces and non-faces. We trace the moving face with Kalman filter, which uses the static information of the detected faces and the dynamic information of changes between previous and current frames. To make a real-time tracking system, we reduce processing time by adjusting the frequency of face detection. In this experiment, the proposed system showed an average tracking rate of 95.5% and processed at 15 frames per second. This means the system is robust enough to track faces in real-time.

• Journal of Broadcast Engineering
• /
• v.19 no.1
• /
• pp.64-74
• /
• 2014

The primary and ambient signal decomposition of a stereo sound is a key step to the stereo upmix. The principal component analysis (PCA) is one of the most widely used methods of primary-ambient signal decomposition. However, previous PCA-based decomposition algorithms assume that stereo sound sources are only amplitude-panned without any consideration of phase difference. So it occurs some performance degradation in case of live recorded stereo sound. In this paper, we propose a new PCA-based stereo decomposition algorithm that can consider the phase difference between the channel signals. The proposed algorithm overcomes limitation of conventional signal model using PCA with phase alignment. The phase alignment is realized by using inter-channel phase difference (IPD) which is widely used in parametric stereo coding. Moreover, Enhanced Modified PCA(EMPCA) is combined to solve the problem of conventional PCA caused by Primary to Ambient energy Ratio(PAR) and panning angle dependency. The simulation results are presented to show the improvements of the proposed algorithm.

• Journal of Electrical Engineering and Technology
• /
• v.12 no.6
• /
• pp.2388-2398
• /
• 2017

In this study, we propose a robust face recognition system to pose variations based on automatic pose estimation. Radial basis function neural network is applied as one of the functional components of the overall face recognition system. The proposed system consists of preprocessing and recognition modules to provide a solution to pose variation and high-dimensional pattern recognition problems. In the preprocessing part, principal component analysis (PCA) and 2-dimensional 2-directional PCA ($(2D)^2$ PCA) are applied. These functional modules are useful in reducing dimensionality of the feature space. The proposed RBFNNs architecture consists of three functional modules such as condition, conclusion and inference phase realized in terms of fuzzy "if-then" rules. In the condition phase of fuzzy rules, the input space is partitioned with the use of fuzzy clustering realized by the Fuzzy C-Means (FCM) algorithm. In conclusion phase of rules, the connections (weights) are realized through four types of polynomials such as constant, linear, quadratic and modified quadratic. The coefficients of the RBFNNs model are obtained by fuzzy inference method constituting the inference phase of fuzzy rules. The essential design parameters (such as the number of nodes, and fuzzification coefficient) of the networks are optimized with the aid of Particle Swarm Optimization (PSO). Experimental results completed on standard face database -Honda/UCSD, Cambridge Head pose, and IC&CI databases demonstrate the effectiveness and efficiency of face recognition system compared with other studies.