• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.14 no.3
• /
• 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.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.15 no.4
• /
• pp.14-19
• /
• 1978

Here a method is shown to extract features from two multi-variable classes by using the coordinate systems transformed by one-class and mixture normalization algorithms. Some properties and implemented results of this technique are described. Also, comparision of these features with factor analysis results is performed. This method is thought to be more powerful one, in feature extraction sense, than factor analysis.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.4 s.304
• /
• pp.59-68
• /
• 2005

This paper proposes a face recognition under varying poses using local area obtained by side-view pose normalization. General normalization methods for face recognition under varying pose have a problem with the information about invisible area of face. Generally this problem is solved by compensation, but there are many cases where the image is distorted or features lost due to compensation .To solve this problem we normalize the face pose in side-view to reduce distortion that happens mainly in areas that have large depth variation. We only use undistorted area, removing the area that has been distorted by normalization. We consider two cases of yaw pose variation and pitch pose variation, and by experiments, we confirm the improvement of recognition performance.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.10
• /
• pp.37-43
• /
• 2021

Using the acoustic features of speech, important social and linguistic information about the speaker can be obtained, and one of the key features is the dialect. A speaker's use of a dialect is a major barrier to interaction with a computer. Dialects can be distinguished at various levels such as phonemes, syllables, words, phrases, and sentences, but it is difficult to distinguish dialects by identifying them one by one. Therefore, in this paper, we propose a lightweight Korean dialect classification model using only MFCC among the features of speech data. We study the optimal method to utilize MFCC features through Korean conversational voice data, and compare the classification performance of five Korean dialects in Gyeonggi/Seoul, Gangwon, Chungcheong, Jeolla, and Gyeongsang in eight machine learning and deep learning classification models. The performance of most classification models was improved by normalizing the MFCC, and the accuracy was improved by 1.07% and F1-score by 2.04% compared to the best performance of the classification model before normalizing the MFCC.

• Proceedings of the IEEK Conference
• /
• 2000.09a
• /
• pp.465-468
• /
• 2000

얼굴 영상은 똑같은 표정의 같은 사람이라도 조명에 따라 매우 다른 얼굴 영상으로 나타난다. 따라서 본 논문에서는 조명 변화에 강인한 얼굴 인식 방법을 제안한다. 제안된 방법은 오프라인 훈련(off-line training)과 온라인 인식(on-line recognition)의 두 부분으로 이루어져 있다. 오프라인 훈련은 PCA(principal component analysis)를 기반으로 한다. 온라인 인식에서는 조명 변화에 대한 보상, 얼굴 특징의 추출, 그리고 인식을 위한 분류 과정의 3 단계로 구성되어 있다. 오프라인 훈련에서는 전체 훈련 얼굴 영상 데이터에 PCA를 적용하여 조명 변화가 최대한 제외된 특징 벡터 공간을 생성한다. 실제 인식 단계에서는 첫 번째로 입력 영상으로 들어온 얼굴 영상에서 조명의 영향을 보상하기 위해 준동형 필터링(homomorphic filtering) 후 밝기 정규화(normalization)를 취한다. 두 번째 단계에서는 입력 데이터의 차원을 줄이고 얼굴 특징 벡터를 구하기 위해 PCA를 수행한다. 마지막 과정으로서 입력 영상의 특징 벡터들과 오프라인에서 미리 구하여진 특징 벡터들의 유사도를 측정하여 얼굴을 인식하게 된다. 실험 결과 제안된 방법은 기존의 Eigenface 방법에 비해 우수한 성능을 나타내었다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.06c
• /
• pp.432-435
• /
• 2007

본 논문은 신원확인 수단으로 부각되어 관심이 높은 홍채인식에 대한 연구이다 홍채인식 시스템의 경우 홍채영역에 따라 각 영상들의 특징 값이 차지하는 비중이 서로 다르게 분포되어 있고 눈썹이나 조명에 의한 잡음으로 인하여 인식성능에 영향을 미친다. 이 경우 기존에 등록되어 인증된 사용자의 홍채영상일지라도 제대로 인식하지 못하거나 인증에 실패할 수 있으며, 실세계에서의 홍채영역 사용이 원활하지 못하게 된다. 그러므로 단일 생체인식 시스템에서 홍채인식을 할 경우, 중요한 특징을 그대로 유지하고 인식성능을 향상시키기 위해서 획득된 홍채 영상의 정규화와 전처리 과정을 거친 다음 홍채영역을 분할한 후 각 영역에서의 보정치 적용을 통한 특징추출 방법을 제안한다. 또한 웨이블릿 변환과 주성분 분석을 이용하여 인식 성능이 개선된 특징추출 방법임을 보인다.

• Journal of Korea Multimedia Society
• /
• v.11 no.10
• /
• pp.1325-1337
• /
• 2008

The difference of environments between training and recognition is the major reason of degradation of speech recognition. To solve this mismatch of environments, various noise processing methods have been studied. Among them, ERN(log-Energy dynamic Range Normalization) and SEN(Silence Energy Normalization) for normalization of log energy features show better performance than others. However, these methods have a problem that they can hardly achieve normalization for the relatively higher values of log energy features and the environmental mismatch caused by this problem becomes bigger especially in low SNR environments. To solve these problems, we propose applying ARMA filter as post-processing for smoothing log energy features by calculating the moving average in auto-regression scheme. From the recognition results conducted on Aurora 2.0 DB, the proposed method shows improved recognition results comparing with conventional methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.9
• /
• pp.1029-1039
• /
• 1992

This paper presents the filtering method which is processed on the frequency domain among Hangul character recognition methods. It is processed the Hangul character parrern with spatial positioning modulation and mapped the Hangul character element which have spatial position variant feature onto frequency domain, at this time, normalized spatial position and so normalized the character size in frequency domain. And it is grouped the Hangul character element according to the generating position and set the standard pattern, and used each standard character element pattern with character element filter and filtering the character pattern of Hangul character, it is derived the normalized cross correlation function and the coherence function led to the filtering results, and calculated classification threshold.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.6
• /
• pp.834-839
• /
• 2009

It is a key element that the problem of emotional feature extraction based on facial image to recognize a human emotion status. In this paper, we propose an Advanced AAM that is improved version of proposed Facial Expression Recognition Systems based on Bayesian Network by using FACS and AAM. This is a study about the most efficient method of optimal facial feature area for human emotion recognition about random user based on generalized HCI system environments. In order to perform such processes, we use a Statistical Shape Analysis at the normalized input image by using Advanced AAM and FACS as a facial expression and emotion status analysis program. And we study about the automatical emotional feature extraction about random user.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.808-810
• /
• 2005

분 논문에서는 워터마크가 삽입된 영상의 비가시성과 강인성을 보장하기 위하여 주파수 영역 기반인 FFT(Fast Fourier Transform)을 이용하였다. 그리고 영상에 삽입된 워터마크를 정확하게 추출하기 위하여 워터마크에 삽입하는 키 사이의 직교성을 유지할 수 있는 그람-슈미트 정규직교화를 이용하였다. 실험을 통해 살펴본 결과 영상의 특징에 관계없이 랜덤계열에 민감한 워터마크를 추출할 수 있는 정확성 및 신뢰성을 가짐을 알 수 있었다.