• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.373-376
• /
• 2011

얼굴 영상에서 구성요소(눈썹, 눈, 코, 입 등)의 존재에 따라 보는 사람의 얼굴 인식 정확도는 큰 영향을 받는다. 이는 인간의 뇌에서 얼굴 정보를 처리하는 과정은 얼굴 전체 영역 뿐만 아니라, 부분적인 얼굴 구성요소의 특징들도 고려함을 말한다. 비음수 행렬 분해(NMF: Non-negative Matrix Factorization)는 이러한 얼굴 영역에서 부분적인 특징들을 잘 표현하는 기저영상들을 찾아내는데 효과적임을 보여주었으나, 각 기저영상들의 중요도는 알 수 없었다. 본 논문에서는 NMF로 찾아진 기저영상들에 대응되는 인코딩 정보를 SLR(Sparse Logistic Regression)을 이용하여 성별 인식에 중요한 부분 영역들을 찾고자 한다. 실험에서는 주성분분석(PCA)과 비교를 통해 NMF를 이용한 기저영상 및 특징 벡터 추출이 좋은 성능을 보여주고, 대표적 이진 분류 알고리즘인 SVM(Support Vector Machine)과 비교를 통해 SLR을 이용한 특징 벡터 선택이 나은 성능을 보여줌을 확인하였다. 또한 SLR로 확인된 각 기저영상에 대한 가중치를 통하여 인식 과정에서 중요한 얼굴 영역들을 확인할 수 있다.

• Annual Conference on Human and Language Technology
• /
• 2009.10a
• /
• pp.35-39
• /
• 2009

이메일의 사용증가로 수신 메일을 효율적이면서 정확하게 분류할 필요성이 점차 증가하고 있다. 현재의 이메일 분류는 베이지안, 규칙 기반 등을 이용하여 스팸 메일을 필터링하기 위한 이원 분류가 주를 이루고 있다. 클러스터링을 이용한 다원 분류 방법은 분류의 정확도가 떨어지는 단점이 있다. 본 논문에서는 비음수 행렬 분해(NMF, Non-negative Matrix Factrazation)를 기반으로 한 자동 분류 주제 생성 방법과 동적 분류 체계(DCH, Dynamic Category Hierachy) 방법을 결합한 새로운 이메일 분류 방법을 제안한다. 이 방법은 수신되는 이메일을 자동으로 분류하여 대량의 메일을 효율적으로 관리할 수 있으며, 분류 결과 사용자의 요구사항을 만족하지 못하면 메일을 동적으로 재분류 하여 분류 정확률을 높일 수 있다.

• ETRI Journal
• /
• v.36 no.1
• /
• pp.167-170
• /
• 2014

In this letter, we propose an unsupervised framework for speech noise reduction based on the recent development of low-rank and sparse matrix decomposition. The proposed framework directly separates the speech signal from noisy speech by decomposing the noisy speech spectrogram into three submatrices: the noise structure matrix, the clean speech structure matrix, and the residual noise matrix. Evaluations on the Noisex-92 dataset show that the proposed method achieves a signal-to-distortion ratio approximately 2.48 dB and 3.23 dB higher than that of the robust principal component analysis method and the non-negative matrix factorization method, respectively, when the input SNR is -5 dB.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.4 s.310
• /
• pp.46-54
• /
• 2006

Recently, the important of a personal identification is increasing according to expansion using each on-line commercial transaction and personal ID-card. Although a personal ID-card embedded RFID(Radio Frequency Identification) tag is gradually increased, the way for a person's identification is deficiency. So we need automatic methods. Because RFID tag is vary small storage capacity of memory, it needs effective feature extraction method to store personal biometrics information. We need new recognition method to compare each feature. In this paper, we studied the face verification system using Hippocampal neuron modeling algorithm which can remodel the hippocampal neuron as a principle of a man's brain in engineering, then it can learn the feature vector of the face images very fast. and construct the optimized feature each image. The system is composed of two parts mainly. One is feature extraction using NMF(Non-negative Matrix Factorization) and LDA(Linear Discriminants Analysis) mixture algorithm and the other is hippocampal neuron modeling and recognition simulation experiments confirm the each recognition rate, that are face changes, pose changes and low-level quality image. The results of experiments, we can compare a feature extraction and learning method proposed in this paper of any other methods, and we can confirm that the proposed method is superior to the existing method.

• 한국정보통신설비학회:학술대회논문집
• /
• 2009.08a
• /
• pp.143-146
• /
• 2009

There are some image hash methods, in the paper four image hash methods have been compared: FJLT (Fast Johnson- Lindenstrauss Transform), SVD (Singular Value Decomposition), NMF (Non-Negative Matrix Factorization), FP (Feature Point). From the compared result, FJLT method can't be used in the online. the search time is very slow because of the KNN algorithm. So FJLT method has been improved in the paper.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10a
• /
• pp.4-6
• /
• 2003

A holistic representation, such as sparse ceding or independent component analysis (ICA), was successfully applied to explain early auditory processing and sound classification. In contrast, Part-based representation is an alternative way of understanding object recognition in brain. In this paper. we employ the non-negative matrix factorization (NMF)［1］which learns parts-based representation for sound classification. Feature extraction methods from spectrogram using NMF are explained. Experimental results show that NMF-based features improve the performance of sound classification over ICA-based features.

• Journal of IKEEE
• /
• v.13 no.2
• /
• pp.159-166
• /
• 2009

So far, many researches have been done to retrieve music information using static classification descriptors such as genre and mood. Since static classification descriptors are based on diverse content-based musical features, they are effective in retrieving similar music in terms of such features. However, human emotion or mood transition triggered by music enables more effective and sophisticated query in music retrieval. So far, few works have been done to evaluate the effect of human mood transition by music. Using formal representation of such mood transitions, we can provide personalized service more effectively in the new applications such as music recommendation. In this paper, we first propose our Emotion State Transition Model (ESTM) for describing human mood transition by music and then describe a music classification and recommendation scheme based on the ESTM. In the experiment, diverse content-based features were extracted from music clips, dimensionally reduced by NMF (Non-negative Matrix Factorization, and classified by SVM (Support Vector Machine). In the performance analysis, we achieved average accuracy 67.54% and maximum accuracy 87.78%.

• Journal of KIISE:Software and Applications
• /
• v.37 no.5
• /
• pp.378-385
• /
• 2010

The explosive increase in the use of email has made to need email classification efficiently and accurately. Current work on the email classification method have mainly been focused on a binary classification that filters out spam-mails. This methods are based on Support Vector Machines, Bayesian classifiers, rule-based classifiers. Such supervised methods, in the sense that the user is required to manually describe the rules and keyword list that is used to recognize the relevant email. Other unsupervised method using clustering techniques for the multi-category classification is created a category labels from a set of incoming messages. In this paper, we propose a new automatic email multi-category classification method using NMF for automatic category label construction method and dynamic category hierarchy method for the reorganization of email messages in the category labels. The proposed method in this paper, a large number of emails are managed efficiently by classifying multi-category email automatically, email messages in their category are reorganized for enhancing accuracy whenever users want to classify all their email messages.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.4
• /
• pp.303-309
• /
• 2016

This paper proposes a novel approach to the problem of audio signal clustering using a stacked autoencoder. The proposed stacked autoencoder learns an efficient representation for the input signal, enables clustering constituent signals with similar characteristics, and therefore the original sources can be separated based on the clustering results. STFT (Short-Time Fourier Transform) is performed to extract time-frequency spectrum, and rectangular windows at all the possible locations are used as input values to the autoencoder. The outputs at the middle, encoding layer, are used to cluster the rectangular windows and the original sources are separated by the Wiener filters derived from the clustering results. Source separation experiments were carried out in comparison to the conventional NMF (Non-negative Matrix Factorization), and the estimated sources by the proposed method well represent the characteristics of the orignal sources as shown in the time-frequency representation.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.12
• /
• pp.2607-2613
• /
• 2012

The results of inherent knowledge based generic summarization are influenced by the composition of sentence in document set. In order to resolve the problem, this papser propses a new generic document summarization which uses clustering of semantic feature of document and coherence of document cluster. The proposed method clusters sentences using semantic feature deriving from NMF(non-negative matrix factorization), which it can classify document topic group because inherent structure of document are well represented by the sentence cluster. In addition, the method can improve the quality of summarization because the importance sentences are extracted by using coherence of sentence cluster and the cluster refinement by re-cluster. The experimental results demonstrate appling the proposed method to generic summarization achieves better performance than generic document summarization methods.