-
原稿種別: 表紙
p.
Cover1-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
原稿種別: 付録等
p.
App1-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
原稿種別: 付録等
p.
i-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
水本 雅晴
原稿種別: 本文
p.
ii-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
原稿種別: 目次
p.
iii-vi
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
原稿種別: 付録等
p.
App2-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
長沢 伸也
原稿種別: 本文
p.
1-8
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
原稿種別: 付録等
p.
App3-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
石桁 正士
原稿種別: 本文
p.
9-16
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
原稿種別: 付録等
p.
App4-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
前田 陽一郎
原稿種別: 本文
p.
17-20
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In this research, we developed the ecological simulator constructed the immune algorithm that is as similar as possible to life. Furthermore, we denned the robustness, that is, the rate for no attack of a desease in case of getting an antigen. And we tried to evaluate the robustness by using the ecological simulator with immune algorithm. We also report results of simulation performed with the ecological simulator.
抄録全体を表示
-
小橋 昌司, 上浦 尚武, 畑 豊
原稿種別: 本文
p.
21-24
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper shows an application of fuzzy information granulation to medical image segmentation for MR angiography image. The method treats medical image as information, and the granules are the anatomical parts such as blood vessel, fat, etc. Our granulation method consists of two stages : (1) Quantum generation and (2) Integration of them to produce granules. Quantum generation is performed with watershed segmentation. Integration process is achievedby estimating the degree for predefined models. By granulating MR angiography image, we obtain the bloodvessel region. In the experimental results, we show the target M IP and shaded surface display images of the segmented region. They are useful for evaluating the lesions such as the areury.
抄録全体を表示
-
佐々木 崇, 上浦 尚武, 畑 豊
原稿種別: 本文
p.
25-26
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper shows a method to segment the menisci regions from MR image by using fuzzy inference. The method is composed of two steps. First, we segment the cartiladge region by thresholding of the intensity. We extract the candidate region of the menisci as the region between the cartilages. Second, we segment the menisci voxels from the candidate region based on fuzzy if-then rules obtained from knowledge of location and intensity. Our method can successfully segment the menisci region from MR data sets. The generated visualizations will help medical doctor to diagnose the menisci with noninvasive.
抄録全体を表示
-
河津 俊幸, 田中 武昌, 神田 重信
原稿種別: 本文
p.
27-28
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
Aim of this study is to consider the accuracy in differentiating between benign odontogenic tumor (ameloblastoma) and cyst (odontogenic keratocyst) on CT image using neural network. In order to determine the weights and the biases of our neural network system, 53 cases (29 ameloblastomas, and 24 odontogenic keratocysts) were used for the learning mode. Another 20 cases (6 ameloblastomas, and 14 odontogenic keratocysts) were evaluated for determining the accuracy in the test mode. The result showed that the diagnostic accuracy was 85 %. It was suggested that the neural network is useful for differentiating these lesions.
抄録全体を表示
-
田中 武昌, 河津 俊幸, 神田 重信
原稿種別: 本文
p.
29-30
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
Recently the digital dental imaging system (DDIS) has been developed and spread widely. The density distribution of DDIS images, however, are insufficient or inadequate. Therefore, the authors considered the density optimization of DDIS images, and applied the fuzzy reasoning and fuzzy c-means clustering algorithm for this purpose. According to our method, density distribution of most DDIS images could be improved appropriately. Moreover, this method could be applied to the dose reduction of the patients. Consequently, it was suggested that the absorbed dose could be decreased 38.3%.
抄録全体を表示
-
浅海 賢一, 喜多村 直
原稿種別: 本文
p.
31-32
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper presents a diagnostic method of a large-scale circulatory system model by applying a structural analysis to the model, where principal causes of an unusual variable are detected for direct or indirect influence processes. We use Coleman's model HUMAN, which consists of 25 physiological modules, over 200 variables and over 100 parameters. A sensitivity analysis serves to find strong-causal relationship between variables in the complex model, and a structural analysis visualizes hierarchical relationship among variables.
抄録全体を表示
-
吉田 香, 加藤 俊一, 矢鳴 虎夫
原稿種別: 本文
p.
33-36
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
People using information system comes to be not only computer specialists but also ordinary people like school kids or housewives, because of high development of internet technology. For example, when the users who has different interests and different tastes would retrieve some information using same keywords, the conventional system answers just same retrieval results to both of the users. It is not enough for users' requirements. We need any adjustable mechanisms which can treat each user's subjective information for human-centered system. In this paper, we show the interactive interface which is a part of human-centered system through image retrieval experiments.
抄録全体を表示
-
升井 一朗, 宇治 寿隆, 本田 武司, 山川 烈
原稿種別: 本文
p.
37-40
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
The purpose of this study was to investigate the correlation of subjective ratings and fuzzy ratings based on cephaloraetric soft tissue analysis ; (1) Which analysis consistently contribute to subjective ratings? (2) Whether OMF surgeons differ from orthodontists in the correlation between subjective ratings and analyses? Materials were lateral cephalogram traces of twenty females. Rators were 21 OMF surgeons and 20 orthodontists. The result of multiple regression analysis showed the form of nasolabial and labiomental region more influenced the subjective evaluation than soft tissue facial convexity in both OMF surgeons and orthodontists.
抄録全体を表示
-
藪内 賢之, 和多田 淳三, 中森 義輝
原稿種別: 本文
p.
41-42
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In this paper, a fuzzy concept is employed to construct a principal component model which can deal with fuzziness, vagueness or possibility of a system, which is named a possibilistic principal component analysis. The principal component analysis is to analyze a possibility of fuzzy numbers. The possibilistic principal component analysis has three formulations according the portions which the possibilities included in fuzzy numbers are embodied : 1) an eigenvalue, 2) an eigenvector and 3) both eigenvalue and eigenvector. In this paper, we discuss about the first formulation 1) that an eigenvalue is employed to deal with fuzziness of data. The possibilistic principal component analysis is employed in this paper to analyze the features of information technology industry. In this analysis, the financial ratio is employed as indices. And we evaluate the possibility of a company activity in information technology industry.
抄録全体を表示
-
市橋 秀友, 長坂 一徳, 三好 哲也, 田村 香奈子
原稿種別: 本文
p.
43-46
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
Fuzzy c-Varieties by Bezdek et al. is an fuzzy clustering method whose fitting prototypes are hyperplanes and the orthogonal distances from points to the planes are minimized. Since each plane corresponds to the space spanned by the principal component vectors, we apply this clustering algorithm to community medicine cost for simultaneous analysis of clustering and principal component. The entropy regularization by Miyamoto et al. is introduced to the Fuzzy c-varieties algorithm.
抄録全体を表示
-
納富 一宏, 斎藤 恵一, 藤本 哲男, 吉田 正徳, 梅津 光生, 石井 博章
原稿種別: 本文
p.
47-48
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In this article, we introduce a database management system for clinical cases on WWW. And also we propose a data searching method with natural language processing. This method has made it possible for all users of the system to search the database for some clinical cases with any keywords (technical terms), sentences, and documents in Japanese. We implemented the system on a Linux (UNIX) based WWW sever as a CGI application. This system has following features : (1) natural language processing (Japanese) for searching, (2) full-text-database, (3) fuzzy inference engine, (4) automatic indexing for appending a new clinical case into the database.
抄録全体を表示
-
斎藤 恵一, 納富 一宏, 藤本 哲男, 吉田 正徳, 梅津 光生, 内山 明彦
原稿種別: 本文
p.
49-50
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper presents a data-grouping method of data management systems. λ-fuzzy measure and Choquet's integral were employed in this method The fuzzy measures and the parameter λ were detected by a genetic algorithm. This method was utilized in our clinical database system on WWW. As a result, the method realized an efficient grouping and searching of clinical cases.
抄録全体を表示
-
平野 英保, 平野 雄, 新井 安芸彦, 中野 正博, 麻生 めぐみ, 田中 良哉
原稿種別: 本文
p.
51-54
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
We have established and proposed the marker search programs for distinguish tumor markers for research, order and diagnosis. The Program (DSCAN : direct scan) was created for line searching in basically, and possible to be searched by paragraphically. The constructions of data were arranged extremely specific (3), very specific (2), specific (1), not important (0.5) for diagnosis according to the detectable frequencies in organ levels. Specific Score was fuzzy inferred by multiplied frequency (%), order of frequency (1-11), and specificity (0.5, 1, 2, 3). The origins of tumor marker naming were also described with shortened abstracts.
抄録全体を表示
-
中野 正博, 松浦 弘幸, 野田 信男
原稿種別: 本文
p.
55-58
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
We have shown a guiding system for selection of courses in university. The system judges the grade of the aptitude of a high school student for 9 courses from his ability and his character. The system consists of three kinds of questions on the favorite subjects and on his personality. From these answers, the system calculates his aptitude score in each course using fuzzy logic and the expert knowledge. The system has the knowledge on the relation between the character and its suitable course. The system serves decision making and has many similar applications.
抄録全体を表示
-
山下 元, 箭内 美智子, 稲井田 次郎, 須田 宏, 竹中 淑子, 佐藤 章
原稿種別: 本文
p.
59-60
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
The sociometry is one of the measurement and evaluation method of a social structure. We could effectively analyse by applying fuzzy graph. Based on the data obtained from some simple questionnaires, we could measure the fuzzy relations among the members in group. We would not only explain the analysis method of the fuzzy sociogram but also illustrate its practical effectiveness with the case study.
抄録全体を表示
-
王 硯玉, 土谷 武士, 水本 雅晴
原稿種別: 本文
p.
61-62
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In This paper, a new learning algorithm is proposed on the basis of a distance-type fuzzy reasoning method. The effectiveness of the proposed learning algorithm is verified by simulation studies.
抄録全体を表示
-
向野 賢治, 升井 一朗, 本田 武司
原稿種別: 本文
p.
63-64
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
MRSA sepsis is still the most fatal disease among various nosocomial infections in our country. To cure patients with MRSA sepsis, rapid diagnosis is essential. However, it takes several days to identify MRSA in a laboratory, while MRSA sepsis causes DIG within a few days resulting in patient death. Thus, vancomycin therapy often starts too late to cure patients. In this study, we tried to create a rapid diagnosis system for MRSA sepsis using fuzzy inference. Clinical data of 5 patients with the disease were applied to the system, then the usefulness was evaluated.
抄録全体を表示
-
北條 暉幸
原稿種別: 本文
p.
65-66
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
A fuzzy inference was applied to analyze microscars, such as striations and pits on the surfaces of teeth, in relation to mastication and threatening of Japanese macaques. The mastication means the behavior by contacts between maxillary and mandibular teeth. In macaques the threatening behavior occurs between the maxillary canines and the mandibular third premolars. In this paper the results of scanning electron microscopy of these microscars on the teeth were reported, and the reasons of their occurrence and their features in relation to their behaviors were inferred using the fuzzy logic.
抄録全体を表示
-
米田 耕二郎, 前田 幹夫, 中山 悌, 村上 周太
原稿種別: 本文
p.
67-70
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
The way to move a wheelchair in the hospital is operation by its crew or its crew's helper usually. Authors assume that the disable person, the old person and the patient use the wheelchair, and suggest the wheelchair had Navigation System on board that it is possible to arrive at a purpose point with the easy operation or semiautomatic operation. We describe the way to search route for navigation, for recognition of object and arrival at objective.
抄録全体を表示
-
野村 行宏, 前田 陽一郎
原稿種別: 本文
p.
71-74
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In this research, as an example of learning that used evolutionary algorithm to a multi-legged robot, make a lag of a phase from a standard locus of each leg that revolves with a regular cycle and a regular locus a gene and it a real number handle can the method that locomotion is caused optimized by causing learnt by using an evolution strategy that the adaptation of self becomes possible in accordance with the environment of surroundings it proposes. Also, it reports regarding the result, because efficacy is verified and carry out a locomotive learn experiment by using walking robot that has 8 legs of real machine.
抄録全体を表示
-
原稿種別: 付録等
p.
App5-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
-
松浦 孝康, 前田 陽一郎
原稿種別: 本文
p.
75-78
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
We propose the method that can avoid deadlocks as a group efficiently, by changing singleton values of fuzzy rule that described the action in an individual agent in accordance with a network of chaotic elements. Furthermore in this paper, we report the result of confirming the efficiency of our proposed algorithm by using the multi-agent robot simulator.
抄録全体を表示
-
中山 悌, 前田 幹夫, 村上 周太
原稿種別: 本文
p.
79-82
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
The patients, old age persons and physical handicapped persons are wheelchair and helpers. When a helper carries one of these persons, he I she must be attend on a person ridden on a wheelchair. The helpers are busy and have many works. Also, other persons with no helper are very tired to navigate their wheelchairs. Therfore, we propose a fuzzy navigation control system for the wheelchair. This system consists of a navigation planning part, a navigation control part and an environment recognition part. In this paper, we report abaut planning method for wheelchair's moving route.
抄録全体を表示
-
米田 耕二郎, 前田 幹夫, 中山 悌, 村上 周太
原稿種別: 本文
p.
83-86
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
The way to move a wheelchair in the hospital is operation by its crew or its crew's helper usually. Authors assume that the disable person, the old person and the patient use the wheelchair, and suggest the wheelchair had Navigation System on board that it is possible to arrive at a purpose point with easy operation or semiautomatic operation. We describe the way to search route for navigation, for recognition of object and arrival at objective.
抄録全体を表示
-
根本 宗記, 市橋 秀友, 三好 哲也
原稿種別: 本文
p.
87-90
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper proposes a batch and supervised type of the Learning Vector Quantization (LVQ), which is an extension of the fuzzy Kohonen Clustering Networks (FKCN) by Tsao, Bezdek and Pal. While they integrated the Fuzzy c-Means (FCM) with the unsupervised type LVQ, we integrate with the supervised type LVQ.
抄録全体を表示
-
豊浦 由浩, 藪内 賢之, 和多田 淳三
原稿種別: 本文
p.
91-94
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In the case where data which come out of several different systems are analyzed by a regression model, we should separate some partition of data which come from one system and analyze each of partitions by a regression model in the proper way. Bezdek has proposed a switching regression model based on cluster analysis and regression analysis. In this study, we propose the genetic model of fuzzy switching regression model based on a fuzzy regression model.
抄録全体を表示
-
宮本 和典, 矢鳴 虎夫, 永松 正博
原稿種別: 本文
p.
95-96
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
A new type of neuron model is proposed including the simulated synaptic function where the symbolized particles are generated, emitted, stored and released towards the neurons of the next layer, as a simplified analogy of the function that real neurons and synapses do by chemical particles. The simulated system of the model was applied to the analysis of voiced signals, and obtained useful results such that even with a single neuron including the unit simulated synapse, the voices examined demonstrated a high degree of recognition. The reaction mechanism in simulated synaptic region would be shown by the mass strategy.
抄録全体を表示
-
越水 孝
原稿種別: 本文
p.
97-100
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
Neural computing has emerged as a practical and powerful tool for "nonlinear" multivariate statistical analysis. In this paper, nonlinear discriminant analysis using a neural network is considered and the probabilistic interpretation of the network output is discussed in a two-class discriminant problem. The principle of the likelihood in network models is employed based on a probabilistic approach regarding the connection weights of the network as unknown parameters. Neural network model may be sensitive to outlying responses. Some diagnostic measures, which had been developed in regression models, are applied in order to detect such observations and quantify their effects in model fitting. Additionally, the information criterion, EIC, and estimates of a true error rate of a prediction rule are discussed when fitting neural network models by employing the bootstrap method.
抄録全体を表示
-
谷川 浩司
原稿種別: 本文
p.
101-102
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
I have investigated the drug interaction checking technology to apply neural network. It is important to check inter-prescription because of the risk on drug interactions. There is some probles in the case of the using current system of the interaction check. The network using this study was the feed forward type of three phases, each drug ID code was coded to the binary code. The backpropagation algorith was used as learning procedure, this computer program was developed using pascal as programming language. The learning performance was utility because absolute error was about 3%.
抄録全体を表示
-
山本 哲哉, 前田 陽一郎
原稿種別: 本文
p.
103-106
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In this research,we proposed a method to improve the search capacity of genetic algorithm. The genetic operations of a simple GA are mainly constructed by the crossover and the mutation. The mutation almost works to hold the variety of solutions in the first stage. The crossover almost works to converge solutions in the last stage. The former play a trick mainly in the last stage and the later in the first stage. Therefore, we improved the search capacity of Simple GA by controlling the mutation rate and the selection rate with fuzzy rules. Simulation results of Traveling Salesman Problem are also shown in this paper.
抄録全体を表示
-
有川 晴彦, 水本 雅晴
原稿種別: 本文
p.
107-108
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
An unexpected non-linearity on fuzzy reasoning output exists in the case of using MIN operation (still the most popular one) and some other operations. The unexpected non-linearity becomes more critical as the number of inputs becomes more. This paper discusses and evaluates it.
抄録全体を表示
-
王 碩玉, 水本 雅晴, 土谷 武士
原稿種別: 本文
p.
109-110
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
In this paper, a new fuzzy reasoning method is proposed based on the electric circuit net theory. Concretely, given fuzzy rules will expressed in equivalence with an electric circuit and the distance of the input and the antecedent part is assumed to be resistance. Reasoning by the current of branch i s an idea of this paper. The proposed fuzzy reasoning method having the following advantages. (1) A high-speed reasoning can be achieved with simple electric circuit. (2) It is possible to apply also for multi-fold and multi-stage fuzzy rules. (3) It is easy to design fuzzy rules. (4) Reasoning consequence also becomes convex fuzzy set if fuzzy sets are convex for the consequent part of the conditional proposition. (5) Modus ponens is strictly satisfied. (6) It is possible to apply for sparse fuzzy rule bases.
抄録全体を表示
-
水本 雅晴
原稿種別: 本文
p.
111-114
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper proposes an additive fuzzy functional reasoning method which aggregates a final inference result by using sum operation rather than max from inference results obtained by fuzzy rules whose consequent part consits of a fuzzy function. The additive fuzzy functional reasoning method is shown to be reduced to a functional reasoning method with a compatibility function which represents the characteristics such as area, base length and importantness of the fuzzy functional rule. The functional reasoning method with compatibility function includes ordinary functional reasoning method by Sugeno and weighted functional reasoning method by Mizumoto as special cases.
抄録全体を表示
-
野田 信雄, 中野 正博, 松浦 弘幸
原稿種別: 本文
p.
115-116
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
Our purpose is to study a stream of blood around the valve in mechanical heart. It is so difficult to solve the problem analytically. We investigate the property of stationary stream around the valve in the tube and describe how to solve this problem using the numerical calculations.
抄録全体を表示
-
平野 英保, 平野 雄, 新井 安芸彦
原稿種別: 本文
p.
117-120
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
We have established how to regulate tissue histology by using Fuzzy Set Chaos Theory (BMFSA Kyushu). We applied this theory on to metastatic tumor cell that proliferate rapidly and migrate into suburban organ and tissue. The basic regulation was to plane gene expression by programming with F-BASIC/WINDOWS to envelop metastatic malignant cell by commanding fibroblast to move around the metastatic cells in the tissue. The ON commands we used were 1) synthesize collagens by ascorbic acid, 2) increase collagen mRNA translation by adding glycine, proline, cysteine, 3) envelop metastatic cell by combining collagens with elastin activated by calcium, 4) proliferate fibrogenic cells by thiamin, riboflavin, pyridoxine, cyanocobalamine. A supplement stop was used for the OFF command. Commands were transduced by blood current, and monitored by X-ray photography and CT-scan.
抄録全体を表示
-
吉川 歩
原稿種別: 本文
p.
121-122
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
This paper aims to reexamine two kinds of mutuality between fuzzy concepts, i.e. "similarity" and "matching" through three psychological experiments. Based on the results from this study, the findings on the mutuality in Yoshikawa (1998) are reviewed. Therefore, the correlation analysis between 30 indices and subjective ratings of the mutuality confirms the following. The indices related to distance along element axis correspond to the subjective ratings than those that express degree of overlap. The other findings, however, are not verified from these results.
抄録全体を表示
-
山川 あす香, 市橋 秀友, 三宅 哲也
原稿種別: 本文
p.
123-126
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
We formulate a new correspondence analysis based on the L_s-norm and propose a neural solution algorithm by the interior point method of mathematical programming. We analyze the psychological feelings for senior-simulations which are for the mimetic experience of advanced aged persons. We show that the psychological feelings or the attitude to the senior-simulation vary by the personalities and interests of youngsters and are clearly classified by our proposed method.
抄録全体を表示
-
斉藤 梨絵, 渡辺 秀彦, 矢鳴 虎夫
原稿種別: 本文
p.
127-129
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー
To grope for better interface or new communication form with machines, like a computer, studies on face and look have been done in various kinds of field. Then using characteristic point data extracted from man face, the similarity of each looks such as joy or angry is claimed and which look is more similar to the optional look is judged. And sending out a questionnaire on ten people to know man's general idea of look, we could find from the result that the importance of each characteristic point become characteristic proportion on each look.
抄録全体を表示
-
白濱 成希, 矢鳴 虎夫
原稿種別: 本文
p.
130-131
発行日: 1998年
公開日: 2017/10/25
会議録・要旨集
フリー
It is important to make a computer that can deal with human sensibility. But this is difficult problem by the reason that it includes subjectivity, vagueness, ambiguity and conditional dependence. For example, the more we try to construct a general database that can express human's sensibility, the more the database loses its information. It is important that these systems include an adjustable mechanism that reflects subjective response of each person. And values of each attribute in the database or parameter of the system should adjust to personality. This paper proposes a best adjustable system to person so as to deal with human's sensibility on computer.
抄録全体を表示
-
原稿種別: 索引
p.
132-
発行日: 1998/11/07
公開日: 2017/10/25
会議録・要旨集
フリー