• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.6
• /
• pp.551-557
• /
• 2013

In intelligent robot research, hand gesture recognition has been an important issue. And techniques that recognize simple gestures are commercialized in smart phone, smart TV for swiping screen or volume control. For gesture recognition, robust hand detection is important and necessary but it is challenging because hand shape is complex and hard to be detected in cluttered background, variant illumination. In this paper, we propose efficient hand detection algorithm for detecting pointing hand for recognition of place where user pointed. To minimize false detections, ROIs are generated within the compact search region using skin color detection result. The ROIs are verified by HOG-SVM and pointing direction is computed by both detection results of head-shoulder and hand. In experiment, it is shown that proposed method shows good performance for hand detection.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.36 no.10
• /
• pp.1171-1179
• /
• 2012

In this study, we consider a dynamic positioning system (DPS) design problem that can be extended to many application fields. Toward this end, tracking and positioning control problems are discussed. In particular, we design a tracking control system that incorporates an observer based on the 2-DOF servo system design approach in order to obtain the desired state information. In the case of observer design, a weighted $H_{\infty}$ error bound approach for a state estimator is considered. Based on an algebraic Riccati equation (inequality) approach, a necessary and sufficient condition for the existence of a full-order estimator that satisfies the weighted $H_{\infty}$ error bound is introduced. The condition for the existence of the estimator is denoted by a linear matrix inequality (LMI) that yields an optimized solution and the observer gain.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.4
• /
• pp.210-216
• /
• 2005

We Propose a new feature normalization scheme based on eigenspace for achieving robust speech recognition. In general, mean and variance normalization (MVN) is Performed in cepstral domain. However, another MVN approach using eigenspace was recently introduced. in that the eigenspace normalization Procedure Performs normalization in a single eigenspace. This Procedure consists of linear PCA matrix feature transformation followed by mean and variance normalization of the transformed cepstral feature. In this method. 39 dimensional feature distribution is represented using only a single eigenspace. However it is observed to be insufficient to represent all data distribution using only a sin91e eigenvector. For more specific representation. we apply unique na independent eigenspaces to cepstra, delta and delta-delta cepstra respectively in this Paper. We also normalize training data in eigenspace and get the model from the normalized training data. Finally. a feature space rotation procedure is introduced to reduce the mismatch of training and test data distribution in noisy condition. As a result, we obtained a substantial recognition improvement over the basic eigenspace normalization.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.38 no.3
• /
• pp.1-12
• /
• 2001

We consider the robust control problem for non-minimum phase(NMP) systems with parametric uncertainty. First, a new method that translates such an uncertain NMP system into a interval family of minimum phase(MP) transfer functions followed a time delay term in the form of Pade' approximation is presented. The controller to be proposed consists of a compensator with Smith predictor structure, so that it can compensate the time delay behaviour due to NMP plant. Therein, the main feedback controller for a family of MP plants has been designed by using quantitative feedback theory(QFT) such that satisfies the robust stability against the structured uncertainty. The stability and performance of overall system are examined through an illustrative example.

• The Journal of the Acoustical Society of Korea
• /
• v.36 no.1
• /
• pp.23-29
• /
• 2017

A concept of a VSA (Virtual Source Array) is the method for an acoustic spatio-temporal focus at a selected location in the outbound direction with respect to the VSA without the need of a probe source as combines a TRP (Time-Reversal Processing) and time-delay and beam-steering. However, in TRP using the VSA concept, it is limited to the critical angle and the short distances relevant to the VSA. In this paper, the waveguide invariant theory is applied to the VSA concept to refocus the received field at ranges greater other than the critical angle and the short ranges by shifting the focused field. The suggested method is verified via numerical simulation, and the results show that the robust acoustic focusing is achieved on the selected location regardless of the limitation on the conventional VSA concept.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.6
• /
• pp.505-510
• /
• 2016

This paper presents a searching method of corresponding points robust to rotational error for scan-matching used for SLAM(Simultaneous Localization and Mapping) in mobile robot. A differential driving mechanism is one of the most popular type for mobile robot. For driving curved path, this type controls the velocities of each two wheels independently. This case increases a wheel slip of the mobile robot more than the case of straight path driving. And this is the reason of a drifting problem. To handle this problem and improves the performance of scan-matching, this paper proposes a searching method of corresponding points using extraction of a closest point based on rotational radius of the mobile robot. To verify the proposed method, the experiment was conducted using LRF(Laser Range Finder). Then the proposed method is compared with an existing method, which is an existing method based on euclidian closest point. The result of our study reflects that the proposed method can improve the performance of searching corresponding points.

• The Journal of Korean Institute of Information Technology
• /
• v.17 no.5
• /
• pp.13-20
• /
• 2019

Recently, object detection technology based on CNN has been actively studied. Object detection technology is used as an important technology in autonomous vehicles, intelligent image analysis, and so on. In this paper, we propose a rotation change robust object detection system by applying MI-FL (Moment Invariant-Feature Layer) to SSD (Single Shot Multibox Detector) which is one of CNN-based object detectors. First, the features of the input image are extracted based on the VGG network. Then, a total of six feature layers are applied to generate bounding boxes by predicting the location and type of object. We then use the NMS algorithm to get the bounding box that is the most likely object. Once an object bounding box has been determined, the invariant moment feature of the corresponding region is extracted using MI-FL, and stored and learned in advance. In the detection process, it is possible to detect the rotated image more robust than the conventional method by using the previously stored moment invariant feature information. The performance improvement of about 4 ~ 5% was confirmed by comparing SSD with existing SSD and MI-FL.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.600-604
• /
• 2018

한국어 문장 분류는 주어진 문장의 내용에 따라 사전에 정의된 유한한 범주로 할당하는 과업이다. 그런데 분류 대상 문장이 띄어쓰기 오류를 포함하고 있을 경우 이는 분류 모델의 성능을 악화시킬 수 있다. 이에 한국어 텍스트 혹은 음성 발화 기반의 문장을 대상으로 분류 작업을 수행할 경우 띄어쓰기 오류로 인해 발생할 수 있는 분류 모델의 성능 저하 문제를 해결해 보고자 문장 압축 기반 학습 방식을 사용하였다. 학습된 모델의 성능을 한국어 영화 리뷰 데이터셋을 대상으로 실험한 결과 본 논문이 제안하는 문장 압축 기반 학습 방식이 baseline 모델에 비해 띄어쓰기 오류에 강건한 분류 성능을 보이는 것을 확인하였다.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2006.06a
• /
• pp.63-66
• /
• 2006

음성인식시스템은 사람이 별도의 장비 없이 음성만으로 시스템의 사용이 가능한 편리한 장점을 지니고 있으나 여러 가지 기술적인 어려움과 실제 환경의 낮은 인식률로 폭넓게 사용되지 못한 상황이다. 그 중 배경잡음은 음성인식의 인식률을 저하시키는 원인으로 지적 받고 있다. 이러한 잡음환경에 있는 ASR(Automatic Speech Recognition)의 성능 향상을 위해 외측억제 기능 이 추가된 파라미터를 제안한다. ASR 에서 널리 사용되는 파라미터인 MFCC을 본 논문에서 제안한 파라미터와 HMM를 이용하여 인식률을 비교하여 성능을 비교하였다. 실험결과를 통해 제안된 파라미터의 사용을 통해 잡음환경에 있는 ASR의 성능 향상을 확인할 수 있었다.

• Annual Conference on Human and Language Technology
• /
• 2008.10a
• /
• pp.69-74
• /
• 2008

본 논문은 인용필드 정규화와 인용매칭의 관계에 대한 분석을 제시한다. 인용매칭은 논문에서 수집된 인용레코드의 인용필드들 간의 비교 결과를 조합하여 동일 논문의 참조여부를 판별하여 인용레코드를 군집화한다. 따라서 인용매칭에 성능을 높일 수 있는 인용필드와 인용매칭 성능의 관계에 대한 연구가 필요하다. 본 논문에서는 인용필드 정규화 및 필드 별 결합에 의하여 인용매칭 성능이 변화하는 것을 보였다. 또한, 인용매칭 성능을 인용필드 유사도와의 관점에서 분석하였다. 앞으로, 인용필드 정규화 및 특성이 인용매칭에 미치는 영향에 대한 이해를 넓혀, 이를 인용매칭에 활용할 수 있으리라 여겨진다.