• The Journal of the Acoustical Society of Korea
• /
• 제17권3E호
• /
• pp.12-21
• /
• 1998

In this paper, we describe the one channel five-way, V/U/M/N/S (Voice/Unvoice/Nasal/Silent), classification algorithm for automatically classifying speech. The decision making process is viewed as a pattern viewed as a pattern recognition problem. Two aspects of the algorithm are developed: feature selection and classifier type. The feature selection procedure is studied for identifying a set of features to make V/U/M/N/S classification. The classifiers used are a vector quantization (VQ), a neural network(NN), and a decision tree method. Actual five sentences spoken by six speakers, three male and three female, are tested with proposed classifiers. From a set of measurement tests, the proposed classifiers show fairly good accuracy for V/U/M/N/S decision.

• 제어로봇시스템학회논문지
• /
• 제12권3호
• /
• pp.284-288
• /
• 2006

In this paper, we apply several pattern recognition algorithms to emotion recognition system with speech signal and compare the results. Firstly, we need emotional speech databases. Also, speech features for emotion recognition is determined on the database analysis step. Secondly, recognition algorithms are applied to these speech features. The algorithms we try are artificial neural network, Bayesian learning, Principal Component Analysis, LBG algorithm. Thereafter, the performance gap of these methods is presented on the experiment result section. Truly, emotion recognition technique is not mature. That is, the emotion feature selection, relevant classification method selection, all these problems are disputable. So, we wish this paper to be a reference for the disputes.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2006년도 춘계학술대회 학술발표 논문집 제16권 제1호
• /
• pp.347-350
• /
• 2006

In this paper, we apply several pattern recognition algorithms to emotion recognition system with speech signal and compare the results. Firstly, we need emotional speech databases. Also, speech features for emotion recognition is determined on the database analysis step. Secondly, recognition algorithms are applied to these speech features. The algorithms we try are artificial neural network, Bayesian learning, Principal Component Analysis, LBG algorithm. Thereafter, the performance gap of these methods is presented on the experiment result section. Truly, emotion recognition technique is not mature. That is, the emotion feature selection, relevant classification method selection, all these problems are disputable. So, we wish this paper to be a reference for the disputes.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2008년도 하계종합학술대회
• /
• pp.903-904
• /
• 2008

This paper suggests a simple scheme of block motion estimation in which the search pattern selection is based on the classification of motion content available in the spatio temporal neighboring blocks. The search area is divided into eight sectors and the search pattern selection is also based on the direction of predicted motion vector. Experimental results show that the proposed algorithm has achieved good predicted image quality measured in terms of PSNR and has very less computational complexity.

• 한국기계가공학회지
• /
• 제11권5호
• /
• pp.29-34
• /
• 2012

This study describes the selection of optimum machining conditions for a high-hardness resin by using a large 5-axis machine. The experiments were conducted to examine the main factors that affect the surface roughness, such as the spindle speed, axial and radial depths of the cut, and pattern of the cutter path. To analyze the experiment results, the factor with the biggest impact on machining was determined using the smaller-the-better characteristic of the Taguchi method; the effectiveness of the experiment was then confirmed by verifying the selected optimum machining condition.

• 한국정보처리학회논문지
• /
• 제3권3호
• /
• pp.429-438
• /
• 1996

본 논문에서는 주어진 문제를 해결하기 위해 사용될 최적의 다충 퍼센트론을 구성 하기 위한 하나의 점진적 학습 방법을 제시한다. 고정된 크기의 트레이닝 패턴 집합을 반복적으로 사용하는 기존의 알고리즘들과는 달리, 제시되는 방법에서는 학습 패턴의 수를 점차 증가시키면서 전체 데이터를 학습하기 위해 필요하고도 충분한 은닉뉴런의 수를 찾는다. 이와 같이 신경망 크기의 최적화에 학습 패턴을 점차적으로 선택하여 늘려나감으로써 일반화 능력과 학습 속도가 기존의 방법에서보다 향상됨을 필기체 숫자인식 문제에 있어서 실험적으로 보여준다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2006년도 추계학술대회 학술발표 논문집 제16권 제2호
• /
• pp.179-182
• /
• 2006

현재 감정을 인식할 수 있는 방법으로는 음성, 뇌파, 심박, 표정 등 많은 방법들이 존재한다. 본 논문은 이러한 방법 중 음성 신호를 이용한 방법으로써 특징들은 크게 피치, 에너지, 포만트 3가지 특징 점을 고려하였으며 이렇게 다양한 특징들을 사용하는 이유는 아직 획기적인 특징점이 정립되지 않았기 때문이며 이러한 선택의 문제를 해결하기 위해 본 논문에서는 특징 선택 방법 중 Multi Feature Selection(MFS) 방법을 사용하였으며 학습 알고리즘은 Self Organizing Map 알고리즘을 이용하여 음성 신호의 감정 특징 패턴을 분류하는 방법을 제안한다.

• 한국축산식품학회지
• /
• 제21권2호
• /
• pp.116-125
• /
• 2001

This survey is carried out in order to understand the perception and eating pattern of chicken further processed product of the housewives in Korea. 894 questionnaires were used for analysis. Chicken is loved by all age groups of people, especially at home of young housewives and of high educational background. The reasons of purchasing chicken further processed products were \"Children(57.8%)\", \"Convenience(43.2%)\" and \"Taste(30.1%)\". The average purchasing frequency of chicken further processed products is \"Once a month\". The criteria for selection of chicken further processed product were \"Hygiene(64.9%)\", \"Expiration date(58.4%)\", \"Brand(25.4%)\", \"Taste(21.9%)\" and \"Price(19.0%)\". Housewives depend on the \"Brand\" in selection of chicken further processed products as the educational background improves. The shopping place of purchasing chicken further processed product were \"Discount store(43.1%)\", Department store(27.4%)\" and so on. The pattern of consumption \"Buy chicken further processed products and cook at home(5.9%)\" was still very low. \"Order for delivery(57.4%)\" increased rapidly for chicken consumption. The housewives suggest to produce chicken further processed products \"Hygienically(16.7%)\", \"Improve taste(12.2%)\" and \"Restrict food additives(11.9%)\" among 311 responses.

• 한국가금학회지
• /
• 제21권2호
• /
• pp.101-111
• /
• 1994

From the observations of dietary self-selection by growing pullets, step-up protein or reverse protein and single-stage low protein pullet feeding systems were developed. They offered another pullet feeding concept that appears to control the body weight effectively and to reduce the consumption of feed and nutrients without impairment of subsequent laying performance. It is obvious from the feed and nutrient consumption pattern of layers fed diets for self-selection of energy, protein and calcium that they have a daily cyclic requirement rather than a constant requirement for nutrients. It seems that a practical self-selective feeding system is needed to meet the daily cyclic requirement for nutrients without consuming an excess of energy and protein at certain times of the day as compared to the complete or single diet where layers have to consume extra energy and protein in the afternoon when they have a specific appetite mainly for calcium.

• 정보통신설비학회논문지
• /
• 제8권1호
• /
• pp.18-23
• /
• 2009

The conventional channel selection method for a ZigBee communication network basically detects the energy values in all channels. In the ISM band, no license is required to use channels in this band, so there may exist various interference factors in this band. It is well known that WLAN is the major interference factor degrading the performance of ZigBee in the ISM band. In this paper, we propose an efficient channel selection algorithm which uses the pattern of two types of WLAN channel uses as the a priori information of IEEE 802.11 band IEEE 802.11g. By using the proposed algorithm, we may save the time required to select channels for the ZigBee communications.