• The KIPS Transactions:PartB
• /
• v.14B no.2
• /
• pp.135-140
• /
• 2007

Feature selection is a preprocessing technique commonly used on high dimensional data. Feature selection studies how to select a subset or list of attributes that are used to construct models describing data. Feature selection methods attempt to explore data's intrinsic properties by employing statistics or information theory. The recent developments have involved approaches like correlation method, dimensionality reduction and mutual information technique. This feature selection have become the focus of much research in areas of applications with massive and complex data sets. In this paper, we provide a feature selection method considering data characteristics and generalization capability. It provides a computational approach for feature selection based on fuzzy cluster analysis of its attribute values and its performance measures. And we apply it to the system for classifying computer virus and compared with heuristic method using the contrast concept. Experimental result shows the proposed approach can give a feature ranking, select the features, and improve the system performance.

• Journal of Internet Computing and Services
• /
• v.7 no.2
• /
• pp.23-35
• /
• 2006

This work presents o novel method which extract facial region und features from motion picture automatically and controls the 3D facial expression in real time. To txtract facial region and facial feature points from each color frame of motion pictures a new nonparametric skin color model is proposed rather than using parametric skin color model. Conventionally used parametric skin color models, which presents facial distribution as gaussian-type, have lack of robustness for varying lighting conditions. Thus it needs additional work to extract exact facial region from face images. To resolve the limitation of current skin color model, we exploit the Hue-Tint chrominance components and represent the skin chrominance distribution as a linear function, which can reduce error for detecting facial region. Moreover, the minimal facial feature positions detected by the proposed skin model are adjusted by using edge information of the detected facial region along with the proportions of the face. To produce the realistic facial expression, we adopt Water's linear muscle model and apply the extended version of Water's muscles to variation of the facial features of the 3D face. The experiments show that the proposed approach efficiently detects facial feature points and naturally controls the facial expression of the 3D face model.

• Journal of Navigation and Port Research
• /
• v.46 no.3
• /
• pp.280-288
• /
• 2022

This study examined the diagnostics of abnormalities and faults of equipment, whose rotational speed changes even during regular operation. The purpose of this study was to suggest a procedure that can properly apply machine learning to the time series data, comprising non-stationary characteristics as the rotational speed changes. Anomaly and fault diagnosis was performed using machine learning: k-Nearest Neighbor (k-NN), Support Vector Machine (SVM), and Random Forest. To compare the diagnostic accuracy, an autoencoder was used for anomaly detection and a convolution based Conv1D was additionally used for fault diagnosis. Feature vectors comprising statistical and frequency attributes were extracted, and normalization & dimensional reduction were applied to the extracted feature vectors. Changes in the diagnostic accuracy of machine learning according to feature selection, normalization, and dimensional reduction are explained. The hyperparameter optimization process and the layered structure are also described for each algorithm. Finally, results show that machine learning can accurately diagnose the failure of a variable-rotation machine under the appropriate feature treatment, although the convolution algorithms have been widely applied to the considered problem.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.2
• /
• pp.469-472
• /
• 2005

This paper propose about effective speech feature parameter that have robust character in effect of noise in realizing speech recognition system. Established MFCC that is the basic parameter used to ASR(Automatic Speech Recognition) and DCTCs that use DCT in basic parameter. Also, proposed delta-Cepstrum and delta-delta-Cepstrum parameter that reconstruct Cepstrum to have information for variation of speech. And compared recognition performance in using HMM. For dimension reduction of each parameter LDA algorithm apply and compared recognition. Results are presented reduced dimension delta-delta-Cepstrum parameter in using LDA recognition performance that improve more than existent parameter in noise environment of various condition.

• Proceedings of the KIEE Conference
• /
• 2015.07a
• /
• pp.1351-1352
• /
• 2015

본 논문에서는 지능형 영상 감시 시스템에서 보행자를 검출하고 추적을 수행하기 위해 은닉층 활성함수에 가우시안 대신 FCM를 사용한 RBFNNs 패턴분류기와 객체 추적 알고리즘인 Mean Shift를 융합한 시뮬레이터를 개발한다. 시뮬레이터는 검출부과 추적부로 나누며, 검출부에서는 입력 영상으로부터 기울기의 방향성을 이용한 HOG(Histogram of Oriented Gradient) 특징을 구하고 빠른 처리속도를 위해 PCA 알고리즘을 통해 차원수를 축소하고 pRBFNNs 패턴분류기를 통해 보행자를 검출 한다. 다음 추적부에서 객체 추적 알고리즘인 Mean Shift를 이용하여 검출된 보행자 추적을 수행한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11b
• /
• pp.841-843
• /
• 2005

비디오 데이터 분석은 감시, 검색, 스포츠 경기 자동 요약 등 많은 분야에서 사용되는 기술이다. 그러나 감시 카메라나 스포츠 경기 비디오와 같이 사람의 영역이 저해상도인 환경에서는 포즈 추정, 모델과의 매칭이 어렵기 때문에 제스처 인식 연구는 많이 이루어지고 있지 못하다. 본 논문에서는 카메라가 Pan/Tilt/Zoom 동작을 하고 사람이 빠르게 움직이는 방송용 테니스 비디오에서, 사람을 추출하고, Curvature Scale Space를 기반으로 한 특징을 추출하여 학습된 포즈 모델과 매칭하는 방법과, 차원의 축소를 통해 일련의 포즈들을 학습된 제스처와 매칭하는 방법을 제안한다. 50개의 방송용 테니스 경기 비디오 장면에 대하여 서브 제스처 추출을 수행한 결과, 서브 포즈에 대하여 모델과 매칭이 잘 되고, 매칭이 되지 않는 포즈를 포함하는 시퀀스에 대해서도 강인한

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.1305-1308
• /
• 2013

본 논문에서는 스마트폰 사용자의 실시간 상황 인식을 위한 효과적인 사운드 분류 시스템을 제안한다. 이 시스템에서는 PCM 형태의 사운드 입력 데이터에 대한 전처리를 통해 고요한 사운드와 화이트 노이즈를 학습 및 분류 단계 이전에 미리 여과함으로써, 계산 자원의 불필요한 소모를 막을 수 있다. 또한 에너지 레벨이 낮아 신호의 패턴을 파악하기 어려운 사운드 데이터는 증폭함으로써, 이들에 대한 분류 성능을 향상시킬 수 있다. 또, 제안하는 사운드 분류 시스템에서는 HMM 분류 모델의 효율적인 학습과 적용을 위해 k-평균 군집화를 이용하여 특징 벡터들에 대한 차원 축소와 이산화를 수행하고, 그 결과를 모아 일정한 길이의 시계열 데이터를 구성하였다. 대학 연구동내 다양한 일상생활 상황들에서 수집한 8가지 유형의 사운드 데이터 집합을 이용하여 성능 분석 실험을 수행하였고, 이를 통해 본 논문에서 제안하는 사운드 분류 시스템의 높은 성능을 확인할 수 있었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.05a
• /
• pp.11-12
• /
• 2008

본 논문은 이차원 영상에 투영된 삼차원 인체의 포즈를 인식하기 위하여 이차원 영상에 투영된 인체의 실루엣 정보를 이용하였다. 인체는 삼차원 공간에서 움직이므로 이차원 영상으로 모든 정보를 알아내기에는 부족한 면이 있다. 따라서 본 논문에서는 인체 포즈의 주시 방향을 결정한 후 인체의 실루엣 영상 Convex-hull 특징점 정보를 이용하여 인체의 삼차원 포즈를 인식하였다. 인체의 포즈는 PCA로 차원을 축소하였으며 Diffusion Distance로 데이터베이스의 포즈모델 중 가장 가까운 모델을 선택하였다.

• Korean Journal of Remote Sensing
• /
• v.34 no.6_1
• /
• pp.1009-1024
• /
• 2018

With the increased use of LiDAR (Light Detection and Ranging) that can obtain over millions of point dataset, methodologies for efficient search and dimensionality reduction for the point cloud became a crucial technique. The existing octree-based "parametric algorithm" has proved its efficiency and contributed as a part of PCL (Point Cloud Library). However, the implementation of the algorithm on GPU (Graphics Processing Unit) is considered very difficult because of structural constraints of the octree implemented in PCL. In this paper, we present a method for the parametric algorithm on GPU environment and implement a projection of the queried points on four directions with an improved noise reduction.

• Journal of IKEEE
• /
• v.24 no.4
• /
• pp.1039-1049
• /
• 2020

Neural decoding is important to recognize the user's intention for controlling a neuro-prosthetic hand. This paper proposes a real-time decoding method for multi-channel peripheral neural activity. Peripheral nerve signals were measured from the median and radial nerves, and motion artifacts were removed based on locally fitted polynomials. Action potentials were then classified using a k-means algorithm. The firing rate of action potentials was extracted as a feature vector and its dimensionality was reduced by a self-organizing feature map. Finally, a multi-layer perceptron was used to classify hand motions. In monkey experiments, all processes were completed within a real-time constrain, and the hand motions were recognized with a high success rate.