• 대한전기학회논문지:시스템및제어부문D
• /
• 제53권4호
• /
• pp.277-285
• /
• 2004

We present a new hybrid stereo matching technique in terms of the co-operation of area-based stereo and feature-based stereo. The core of our technique is that feature matching is carried out by the reference of the disparity evaluated by area-based stereo. Since the reference of the disparity can significantly reduce the number of feature matching combinations, feature matching error can be drastically minimized. One requirement of the disparity to be referenced is that it should be reliable to be used in feature matching. To measure the reliability of the disparity, in this paper, we employ the self-consistency of the disunity Our suggested technique is applied to the detection of 3D line segments by 2D line matching using our hybrid stereo matching, which can be efficiently utilized in the generation of the rooftop model from urban imagery. We carry out the experiments on our hybrid stereo matching scheme. We generate synthetic images by photo-realistic simulation on Avenches data set of Ascona aerial images. Experimental results indicate that the extracted 3D line segments have an average error of 0.5m and verify our proposed scheme. In order to apply our method to the generation of 3D model in urban imagery, we carry out Preliminary experiments for rooftop generation. Since occlusions are occurred around the outlines of buildings, we experimentally suggested multi-image hybrid stereo system, based on the fusion of 3D line segments. In terms of the simple domain-specific 3D grouping scheme, we notice that an accurate 3D rooftop model can be generated. In this context, we expect that an extended 3D grouping scheme using our hybrid technique can be efficiently applied to the construction of 3D models with more general types of building rooftops.

• 정보처리학회논문지B
• /
• 제8B권5호
• /
• pp.573-578
• /
• 2001

This research proposes a generalized local feature structure named "LSG(Local Surface Group) for model-based 3D object recognition". An LSG consists of a surface and its immediately adjacent surface that are simultaneously visible for a given viewpoint. That is, LSG is not a simple feature but a viewpoint-dependent feature structure that contains several attributes such as surface type. color, area, radius, and simultaneously adjacent surface. In addition, we have developed a new method based on Bayesian theory that computes a measure of how distinct an LSG is compared to other LSGs for the purpose of object recognition. We have experimented the proposed methods on an object databaed composed of twenty 3d object. The experimental results show that LSG and the Bayesian computing method can be successfully employed to achieve rapid 3D object recognition.

• 한국산학기술학회논문지
• /
• 제7권4호
• /
• pp.649-655
• /
• 2006

본 논문은 화자 독립의 음성인식을 위한 연구로써, DMS 모델에 의한 DMSVQ(Dynamic Multi-Section Vector Quantization) 코드북과 이중 스펙트럼 특징을 이용한 HMM(Hidden Markov Model) 음성인식 방법을 제안한다. 정적 스펙트럼 특징으로서는 LPC ?S스트럼 계수를 이용하였고, 동적 스펙트럼 특징으로는 LPC ?S스트럼의 회귀계수를 사용하였다. 이들 두개의 스펙트럼 특징들을 각각 VQ 코드북으로 양자화되고, DMS 모델을 이용한 HMM은 입력으로써 정적 스펙트럼 특징과 동적 스펙트럼 특징을 받아드림으로써 모델링된다. 제안된 방법에 의한 인식 실험은 기존의 다양한 인식 방법에 의한 인식 실험들과 비교를 위해 동일한 데이터와 조건 하에서 수행하였다. 실험 결과, 본 연구에서 제안한 방법이 기존의 방법들보다 우수한 방법임을 입증하였다.

• ETRI Journal
• /
• 제43권3호
• /
• pp.549-560
• /
• 2021

Cyberattacks are often difficult to identify with traditional signature-based detection, because attackers continually find ways to bypass the detection methods. Therefore, researchers have introduced artificial intelligence (AI) technology for cybersecurity analysis to detect malicious PowerShell scripts. In this paper, we propose a feature optimization technique for AI-based approaches to enhance the accuracy of malicious PowerShell script detection. We statically analyze the PowerShell script and preprocess it with a method based on the tokens and abstract syntax tree (AST) for feature selection. Here, tokens and AST represent the vocabulary and structure of the PowerShell script, respectively. Performance evaluations with optimized features yield detection rates of 98% in both machine learning (ML) and deep learning (DL) experiments. Among them, the ML model with the 3-gram of selected five tokens and the DL model with experiments based on the AST 3-gram deliver the best performance.

• 한국IT서비스학회지
• /
• 제18권2호
• /
• pp.143-159
• /
• 2019

Machine learning (ML) is a method of fitting given data to a mathematical model to derive insights or to predict. In the age of big data, where the amount of available data increases exponentially due to the development of information technology and smart devices, ML shows high prediction performance due to pattern detection without bias. The feature engineering that generates the features that can explain the problem to be solved in the ML process has a great influence on the performance and its importance is continuously emphasized. Despite this importance, however, it is still considered a difficult task as it requires a thorough understanding of the domain characteristics as well as an understanding of source data and the iterative procedure. Therefore, we propose methods to apply deep learning for solving the complexity and difficulty of feature extraction and improving the performance of ML model. Unlike other techniques, the most common reason for the superior performance of deep learning techniques in complex unstructured data processing is that it is possible to extract features from the source data itself. In order to apply these advantages to the business problems, we propose deep learning based methods that can automatically extract features from transaction data or directly predict and classify target variables. In particular, we applied techniques that show high performance in existing text processing based on the structural similarity between transaction data and text data. And we also verified the suitability of each method according to the characteristics of transaction data. Through our study, it is possible not only to search for the possibility of automated feature extraction but also to obtain a benchmark model that shows a certain level of performance before performing the feature extraction task by a human. In addition, it is expected that it will be able to provide guidelines for choosing a suitable deep learning model based on the business problem and the data characteristics.

• 대한음성학회지:말소리
• /
• 제46호
• /
• pp.37-50
• /
• 2003

In this paper, we evaluate and compare the performance of speech/music discrimination based on various feature parameters and classifiers. As for feature parameters, we consider High Zero Crossing Rate Ratio (HZCRR), Low Short Time Energy Ratio (LSTER), Spectral Flux (SF), Line Spectral Pair (LSP) distance, entropy and dynamism. We also examine three classifiers: k Nearest Neighbor (k-NN), Gaussian Mixure Model (GMM), and Hidden Markov Model (HMM). According to our experiments, LSP distance and phoneme-recognizer-based feature set (entropy and dunamism) show good performance, while performance differences due to different classifiers are not significant. When all the six feature parameters are employed, average speech/music discrimination accuracy up to 96.6% is achieved.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2011년도 한국컴퓨터종합학술대회논문집 Vol.38 No.1(A)
• /
• pp.255-258
• /
• 2011

An application of Gammatone auditory filterbank for emotion recognition from speech is described in this paper. Gammatone filterbank is a bank of Gammatone filters which are used as a preprocessing stage before applying feature extraction methods to get the most relevant features for emotion recognition from speech. In the feature extraction step, the energy value of output signal of each filter is computed and combined with other of all filters to produce a feature vector for the learning step. A feature vector is estimated in a short time period of input speech signal to take the advantage of dependence on time domain. Finally, in the learning step, Hidden Markov Model (HMM) is used to create a model for each emotion class and recognize a particular input emotional speech. In the experiment, feature extraction based on Gammatone filterbank (GTF) shows the better outcomes in comparison with features based on Mel-Frequency Cepstral Coefficient (MFCC) which is a well-known feature extraction for speech recognition as well as emotion recognition from speech.

• 인지과학
• /
• 제14권1호
• /
• pp.11-16
• /
• 2003

본 논문은 Gabor 웨이브렛 변환을 이용하여 무표정을 포함한 표정영상에서 얼굴의 주요 요소들의 경계선을 추출한 후, FCM 군집화 알고리즘을 적용하여 무표정 영상에서 저차원의 대표적인 특징점을 추출한다. 무표정 영상의 특징점들은 표정영상의 특징점들을 추출하기 위한 템플릿으로 사용되어지며, 표정영상의 특징점 추출은 무표정 영상의 특징점과 동적 연결모형을 이용하여 개략적인 정합과 정밀한 정합 과정의 두단계로 이루어진다. 본 논문에서는 Gabor 웨이브렛과 FCM 군집화 알고리즘을 기반으로 동적 연결모형을 이용하여 표정영상에서 특징점들을 자동으로 추출할 수 있음을 제시한다. 본 연구결과는 자동 특징추출을 이용한 차원모형기반 얼굴 표정인식[1]에서 얼굴표정의 특징점을 자동으로 추출하는 데 적용되었다.

• 한국멀티미디어학회논문지
• /
• 제24권8호
• /
• pp.1026-1034
• /
• 2021

Tire wear and defect are important factors for safe driving condition. These defects are generally inspected by some specialized experts or very expensive equipments such as stereo depth camera and depth gauge. In this paper, we propose tire safety vision inspector based on deep neural network (DNN). The status of tire wear is categorized into three: 'safety', 'warning', and 'danger' based on depth of tire tread. We propose an attention mechanism for emphasizing the feature of tread area. The attention-based feature is concatenated to output feature maps of the last convolution layer of ResNet-101 to extract more robust feature. Through experiments, the proposed tire wear classification model improves 1.8% of accuracy compared to the existing ResNet-101 model. For detecting the tire defections, the developed tire defect detection model shows up-to 91% of accuracy using the Mask R-CNN model. From these results, we can see that the suggested models are useful for checking on the safety condition of working tire in real environment.

• 한국공간정보시스템학회 논문지
• /
• 제11권1호
• /
• pp.71-78
• /
• 2009

최근 ITS, 텔레매틱스, 유비쿼터스 등의 도입으로 공간 데이터는 다양한 환경에 응용되거나 활용 분야가 점차 증가하고 있고, 수치지도를 일반인들에게 제공함으로써 공간 데이터에 대한 수요가 급증하고 있다. 기존의 수치지도 관리 시스템은 도엽이라는 일정한 단위로 구분하여 공간 데이터를 관리하고 있기 때문에 데이터의 구축은 용이하지만 객체 단위의 데이터 구축, 관리 및 갱신을 효율적으로 지원하기 어렵다. 따라서 이 논문에서는 이러한 문제를 해결하기 위하여 객체기반의 연속적인 지형지물 표현, 공간 데이터의 객체별 이력관리 및 수시갱신이 가능한 객체기반의 데이터 모델을 제안하였다. 제안 모델에서 객체기반 공간 데이터는 각 지형지물에 UFID를 부여하고 도엽 단위로 구축된 수치지도 데이터의 조인 연산을 통해 연속적인 지형지물을 표현하였다. 아울러 갱신으로 인한 변경 데이터를 이력 DB에 시간간격 단위로 저장, 관리하였으며, 제안된 모델의 효율성을 검증하기 위하여 타당성을 분석하였다.