ITE Technical Report
Online ISSN : 2424-1970
Print ISSN : 1342-6893
ISSN-L : 1342-6893
Search
Advanced search
OR
Browse
Search
34.18
Displaying 1-21 of 21 articles from this issue
Show all abstracts
    • |<
    • <
    • 1
    • >
    • >|
  • [title in Japanese]
    Article type: Cover
    Pages Cover1-
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_Cover1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (18K)
  • [title in Japanese]
    Article type: Index
    Pages Toc1-
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_Toc1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (77K)
  • [title in Japanese]
    Article type: Bibliography
    Pages Misc1-
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_Misc1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (42K)
  • Design of Similarity and Impression Evaluation of Music with Local Characteristic
    Masashi MURAKAMI, Toshikazu KATO
    Article type: Article
    Session ID: ME2010-86/AIT2010-78
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    In this research, we analyze how the sound and music relate to humans from the aspect of Kansei engineering. We analyze what features of the sound humans pay attention and how humans interpret sound. Therefore, we divide the signal processing of sound that humans do into four levels. At the physiological level, processing is done by the auditory characteristic. In this level, humans don't interpret the image of the sound yet. There is no subjectivity for the sound. By using auditory characteristic, we investigate the features which help in the case that sound and music is analyzed. We consider that the processing at early stage of auditory nervous system is to extract the change in power, which is obtained from the segmentation of the sound-signals which is divided by band of the frequency and time interval, and its contrast. We also consider the features obtained by that extractation. Moreover, in the cognitive level, we analyze the correlation of that features with the word of interpretation that humans do subjectively. By these modelings, we develop the method of retrieving the sound and music that having the similarity, or having the image that is expressed by any subjective words.
    View full abstract
    Download PDF (720K)
  • Efficacy of Projection of CG in Live Stage Performances
    Masahito Shiba, Asako Soga, Jonah Salz
    Article type: Article
    Session ID: ME2010-87/AIT2010-79
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_5
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We have developed a support system for effective presentations in live stage performances. This system provides functions of using images and videos created by computers and controlling them in real-time. This system is easy and portable to use in theaters for stage staffs who do not have specialized knowledge about computers. We carried out two live stage performances with the system. We got feedbacks from spectators and the stage director that managing visual images by the system was effective. In this article, the structure of the system and our experience in the two performances are described.
    View full abstract
    Download PDF (992K)
  • Development of three-dimensional displays based on human visual field characteristics
    Sumio YANO
    Article type: Article
    Session ID: ME2010-88/AIT2010-80
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_9
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    In this article, a brief principle and examples of development three-dimensional displays are described. The principle of development of three dimensional displays are based on light field reproduction. In the light field reproduction, the displayed object are supposed in the display space, and the light line from display screen are perceived by viewers as the light line from that object. Numerous of light line are necessary for the making of the field of light line. These field of light line is composed of the group of light line, which is radiated to light and left, and also up and down, on the one point of the display screen. Otherwise, only radiated light line to the horizontal direction is also effective, the displayed three-dimensional images have only horizontal disparity in this case. Integral photography, high density directional images and so on are known as the practical method for development of these three-dimensional displays. Some examples of the developed three-dimensional systems based on this principle in our laboratories are described. There are three kind of three-dimensional displays, one is a cubic-type three-dimensional display, which is used on hand, second is a tabletop-type three-dimensional display, which is watched from viewers at the around of the display, third is a large-screen three-dimensional display, which is watched by the several viewers at the same time. These three-dimensional displayed have been developed based on the human visual field characteristics.
    View full abstract
    Download PDF (1170K)
  • Development of Image Retrieval System Using Multiple Key Images
    Kazuki YASUKAWA, Toshikazu KATO
    Article type: Article
    Session ID: ME2010-89/AIT2010-81
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_15
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    Many image search systems, designed to search on only one sample image, share the following 3 problems: (A) Appropriate search results can be expected only when users have images at hand which clearly possess characteristics they want. (B) The same search results are presented when searched by the same images, even if image characteristics intended by users vary. (C) Characteristics which are difficult to express, such as the general color tone and texture, cannot be defined as search keys. To solve these problems, we developed an algorithm to extract common image characteristics from multiple sample images and to estimate likely searches. We developed 2 prototype image search methods: explicitly highlighting details wanted by users from among multiple sample images presented, and estimating characteristics wanted by users from among multiple sample images presented. We evaluated these systems on a database of 1500 landscape images and obtained good results.
    View full abstract
    Download PDF (1147K)
  • Analysis of Communication between Elderly People in Videophone with Supporter Robot
    Yuichi KOYAMA, Tomoko YONEZAWA, Hirotake YAMAZOE, Shinji ABE, Kenji MA ...
    Article type: Article
    Session ID: ME2010-90/AIT2010-82
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_19
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We have proposed communication supporting system using robot, in order to sustain dialogue between elderly people in video communication. We conducted a field trial with twelve elderly people at institutions for the elderly to confirm the feasibility of the system. The results showed that six of twelve conversations sustained around fifteen minutes at which experimenter stopped the conversations, and the other six conversations lasted for 10.0 minutes. Such contributions of the system as (1) user interaction with robot increased in the last half of the conversation in which concentration of the elderly people decreased; (2) the robot utterance sustained the conversation; were confirmed by the analysis of communication of the elderly people.
    View full abstract
    Download PDF (1079K)
  • A Compensation Method for Hand-drawn Illustration by Moving Average
    Hotaka KAWASE, Mikio SHINYA, Michio SHIRAISHI
    Article type: Article
    Session ID: ME2010-91/AIT2010-83
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_23
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    When we trace illustration using paint tools, it is important to compensate fluctuations due to hand-drawing. This paper proposes an improved compensation method based on the moving average method. The moving average method calculates averages of a fixed number of input points and adopts them as modified input points. However, when the pen speed is too fast or the line curvature is too large, the average range can be too large, which makes compensated lines away from the original trace lines. In the proposed method, we adaptively change the filter size according to the line curvatures and the pen speed. Experiments demonstrated the improvements over the previous method.
    View full abstract
    Download PDF (598K)
  • Influence of Size Reduction upon the Impression of Book Jackets
    Nao UEDA, Atsushi SHIMOJIMA
    Article type: Article
    Session ID: ME2010-92/AIT2010-84
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_27
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    In this study, we were interested in the size reduction of book jacket pictures found in online bookstores, and investigated the conditions under which the size reduction influence our impressions. Our data suggested that the degrees of impression change was not affected by the jackets contains illustrations only or contains photographs, they were affected by whether the jackets contain the images of human beings.
    View full abstract
    Download PDF (710K)
  • Modeling Visual Impression of Portrait Photos
    Shunsuke UESAKA, Kazuki YASUKAWA, Riki SHIGEMATSU, Toshikazu KATO
    Article type: Article
    Session ID: ME2010-93/AIT2010-85
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_31
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We have been researching computational modeling of visual impression process for scenery photos. There seems to be another progress for portrait photos. We have experimented the difference of images for portrait photos with facial images and without them. We found that some images words are sensitive to facial parts, while others are sensitive to scenery parts. We have also tried automatic classification of photos according to their predicted images.
    View full abstract
    Download PDF (862K)
  • The Effects of Change in the Composition on Impressions of Japanese Famous Pictures
    Kenji TSUJIMOTO, Masakatu MURAKAMI
    Article type: Article
    Session ID: ME2010-94/AIT2010-86
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_35
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We tried mathematical comparison analysis from the viewpoint of impression evaluation and eye movement. We compared two versions of the same picture, making changes in the composition. As a result, we understood that the impression evaluation changed considerably with changes in composition. The total area of eye movement was also an important factor.
    View full abstract
    Download PDF (1202K)
  • Confirmation of gripping surface status classification using HMM and tactile sensor array
    Wataru MITO, Hirotake YAMAZOE, Shunsuke YOSHIDA, Masahiro TADA, Masayu ...
    Article type: Article
    Session ID: ME2010-95/AIT2010-87
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_39
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We have designed and fabricated a tactile sensors array with three inclined micro-cantilevers on a single sensor embedded in elastomer. Each sensor element can detect both normal and shear stresses. In this paper, we propose classification method of gripping status using sensor array and machine learning algorithm based on hidden marcov model. By integrating multiple sensor data using machine learning algorithm using hidden marcov model, our method allows to detect partial contact status such as non-contact, three types of partial contact, full-contact. Through experiments, we confirmed that our method achieves better classification ratio than existing methods using single sensor data.
    View full abstract
    Download PDF (966K)
  • Analysis on Sense of Surface Materials Based on Local Brightness Patterns
    Wenhao WANG, Yasuhiro TAKEDA, Toshikazu KATO
    Article type: Article
    Session ID: ME2010-96/AIT2010-88
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_43
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    In this paper, we analyze the impact of surface materials on the reasons. First, we use the old methods, while adding higher-order contrast is the amount of local brightness patterns. Through discriminant analysis, in determining a sense of glossy and translucent experimental results show that the higher-order contrast than histogram skewness more effective.
    View full abstract
    Download PDF (792K)
  • Calm & Modest Interaction Method for Modeling User's Interest through Natural Reaction in Wearable Environments
    Hiroaki SUZUKI, Toshikazu KATO
    Article type: Article
    Session ID: ME2010-97/AIT2010-89
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_47
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    With the development of the mobile information and communication environment, the use rate of the mobile telephone suddenly rises, in the various scenes in the everyday life, information distribution to the user is enabled. By this report, I show the calm & modest information that considered the personal situation in a terminal and suggest technique to model interest because observation measures the natural reaction that the user for it shows.
    View full abstract
    Download PDF (684K)
  • Remarks on Generation of Thumbnail Regions from Panorama Images
    Hiroya KITADA, Kazuhito MURAKAMI
    Article type: Article
    Session ID: ME2010-98/AIT2010-90
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_51
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    This paper proposes a method to extract the most essential and characteristic region, thumbnail image, from a panorama image. The density of feature points and luminance distribution are utilized as an index of the image feature and the similarity evaluation of the region, respectively. Some experimental results show the effectiveness of our proposed method.
    View full abstract
    Download PDF (1019K)
  • A study of recognizing movement in hand gesture interface using image processing
    Yuuji ASADA, Isao NISHIHARA, Shizuo NAKANO
    Article type: Article
    Session ID: ME2010-99/AIT2010-91
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_55
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    Recently, the wearable computing that uses the computer anywhere is generalized. Then, we propose the gesture interface that uses the image processing as a hands-free input interface that doesn't have restriction and can use in moving situation. The skeleton model is made by the center point of the hand area and finger positions to identify the movement of a detailed finger. The relevance rate 82% and the recall rate 87.5% were obtained by the experiment, and the proposed method was shown as effective method to use.
    View full abstract
    Download PDF (829K)
  • Grouping tastes in color by personal preference and lifestyles
    Ayako YOMO, Toshikazu KATO
    Article type: Article
    Session ID: ME2010-100/AIT2010-9
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_59
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    Color coordinates in fashion are based on personal taste on color, design, material, and so on, which may differ for each individual. We have applied cluster analysis and relationship analysis among each user and her favorite colorcoordinate. Then, we found that the sense of lifestyle is a dominant factor rather than personal taste for each factor in fashion.
    View full abstract
    Download PDF (777K)
  • [title in Japanese]
    Article type: Appendix
    Pages App1-
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_App1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (97K)
  • [title in Japanese]
    Article type: Appendix
    Pages App2-
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_App2
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (97K)
  • [title in Japanese]
    Article type: Appendix
    Pages App3-
    Published: May 24, 2010
    Released on J-STAGE: September 21, 2017
    DOIhttps://doi.org/10.11485/itetr.34.18.0_App3
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (97K)
    • |<
    • <
    • 1
    • >
    • >|
feedback
 Top