Hiroyo Ishikawa
| Faculty of Humanities and Social Sciences,Department of Business,Management Information Course | Associate Professor |
| Graduate School of Information Sciences,Major of Information Sciences | Associate Professor |
Last Updated :2025/10/07
■Researcher basic information
■Research activity information
Award
Paper
- MoSalC : Visualizing diversity of book exhibition information : Providing books within a digital content space
Hiroyo Ishikawa; Tokutaro Murakami; Kunitake Kaneko
Records management : journal of the Records Management Society of Japan, Dec. 2019, [Reviewed], [Invited]
In this research, we propose MoSaIC (Museum of Shared and Interactive Cataloguing) system to browse book exhibitions. In this system, “catalogues” are created by modeling diversity of relationships and visualized in 3D computer graphics. We propose two methods for creating “catalogues” by books collected for the exhibition. One is grouping books, and the other is constructing a data tree structure by using Nippon Decimal Classification (NDC). We applied these methods to data from some exhibitions that can be accessed on the Internet. As a result, MoSaIC system can express the relevance of exhibitions including the same book(s), and various facets of the book included in seemingly unrelated themes. Furthermore, in NDC tree structure method, the tendency of subject classification for books collected can be visualized. MoSaIC system can link book exhibitions in digital content world, over the time and the places, and provide chance to encounter books from a diversity of the book exhibitions. - Visualizing the relationships among digital contents by MoSaIC - Using for a research of an altarpiece in Italian Renaissance -
じんもんこん2018論文集, 24 Nov. 2018, [Reviewed]
In this paper, an Italian Renaissance altarpiece research paper is catalogued by MoSaIC(Museum of Shared and Interactive Cataloging) system. Its effectiveness is then considered. MoSaIC is a system of modeling various relationships between digital contents to make catalogues with a directed graph and visualizing the catalogues. Some sections of a paper for Pesaro Altarpiece (Giovanni Bellini) are catalogued as cited images and texts as digital content. As the result, we can see the structure of relationships among materials by visualizing the result catalogues with 3D computer graphics. In addition, cataloguing a paper with MoSaIC leads to an analysis of the structure and to understanding the paper. - Polymorphic cataloguing and interactive 3D visualization for multiple context of digital content: MoSaIC
Hiroyo Ishikawa; Kunitake Kaneko
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Feb. 2017, [Reviewed] - Catalogue: Graph Representation of File Relations for a Globally Distributed Environment
Yamato Miyashita; Hiroyo Ishikawa; Fumio Teraoka; Kunitake Kaneko
30TH ANNUAL ACM SYMPOSIUM ON APPLIED COMPUTING, VOLS I AND II, 2015, [Reviewed] - Polymorphic Cataloguing and Viewing System for Using Digital Archives: MoSaIC-II
Hiroyo Ishikawa; Hideo Saito; Yamato Miyashita; Kunitake Kaneko
PROCEEDINGS OF THE 2014 INTERNATIONAL CONFERENCE ON VIRTUAL SYSTEMS AND MULTIMEDIA (VSMM), Dec. 2014, [Reviewed] - Resource-Aware Rendering for Laser Plasma Scanning 3D Display Devices Using Surface Descriptors
中谷文香; 藤代一成; 石川尋代; 斎藤英雄
日本バーチャルリアリティ学会論文誌, Dec. 2012, [Reviewed]
Laser-Plasma Scanning 3D Display (LPSD) is one of the most advanced volume scanning display devices, and plots a unicursal series of illuminants freely in the air. Though the LPSD device has been repeatedly upgraded, even the latest version is only capable of plotting an object at a rate of 5,000 illuminants per second, purely due to the hardware limitation. If we plotted illuminants at a regular interval with a constant brightness, it might possibly lead to loss of geometry and/or surface texture of the displayed object. This paper, therefore, proposes a software-based approach to ameliorate this problem, which controls the brightness and density of plotted illuminants in accordance with existing surface descriptors, to convey salient features of 3D shapes with even a limited number of illuminants. - Surface representation of 3D object for aerial 3D display
Hiroyo Ishikawa; Hayato Watanabe; Satoshi Aoki; Hideo Saito; Satoru Shimada; Masayuki Kakehata; Yuji Tsukada; Hidei Kimura
STEREOSCOPIC DISPLAYS AND APPLICATIONS XXII, Jan. 2011, [Reviewed] - Interactive Sound Creation System with Depth Camera
Hiroyo Ishikawa; Takeki Ihara; Takuya Ujihara; Hideo Saito
Proceedings of 20th International Conference on Artificial Reality and Telexistence, Dec. 2010, [Reviewed]
In this paper, we propose an interactive sound creation system with a depth camera. When an object enters in the selected area of the depth camera, a sound is created interactively according to the acquired data (depth data). This system decides the three elements (pitch, tone and volume) of the sound by the statistical value of the depth data, position and motion of the objects. This system can create two kinds of sound. One is sound effects, like Theremin, which is created by fluctuating pitch frequency and harmonics structure, the other is creating a sound by mixing four kinds of tone selected from the musical source of a synthesizer. In the latter, when pitch is decided from the pitch included in a musical scale, melodies can be created automatically by the motion of objects. When the selected tone and the musical scale are changed, the system can create various sounds and melodies of another musical style. Additionally, an object does not necessarily have to be a human. Therefore, this system is a new style of virtual musical instrument to create sound effects and melodies unconsciously. - Vision-based Interaction with Aerial 3D Display
Hayato Watanabe; Hiroyo Ishikawa; Satoshi Aoki; Seiji Suzuki; Hideo Saito; Satoru Shimada; Tatsumi Kimura; Masayuki Kakehata; Yuji Tsukada; Hidei Kimura
International Conference on 3D Systems and Applications 2010 (3DSA 2010), May, May 2010, [Reviewed]
This paper presents vision-based interaction for the aerial 3D display that can generate illuminating dots sequence in a volumetric screen by laser scanning device. The 3D contents shown in the volumetric screen can be interactively changed and controlled according to the user’s action, which is captured with cameras. As an example of the interactive system, we developed “Jan-Ken Game (rock paper scissors)” system. - Billboard representation for Laser-Plasma Scanning 3D Display
Hiroyo Ishikawa; Hideo Saito
Proceedings of Asiagraph 2009, Oct. 2009, [Reviewed]
In this paper, we propose a new billboard representation for Laser-Plasma scanning display. The Laser-Plasma 3D display can generate plasma luminous bodies at arbitrary position in midair without any screens. Using this 3D display we can display messages in free space as some kind of billboard that represents sequence of characters formed with dots of plasmas. Dots of plasmas sequentially blink 1000 times a second, by vector scanning. However, there is a restriction on the number of the plasmas which human can perceive simultaneously. In this research, we experiment on the difference of representing characters by changing the density of the plasmas and the drawing way of characters on our system. From some experiment, it can be seen that the way of drawing several times by reducing the density of points is stable to recognize a character, even though the drawing time is the same. Furthermore, the character string is considered, and it can be seen that the way of displaying several characters simultaneously is an appropriate method. - Object representation using point sequence for laser plasma 3D display
Hiroyo Ishikawa; Hideo Saitot
Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, May 2009, [Reviewed] - Closed-Line Based Representation of 3D Shape for Point Cloud for Laser-Plasma Scanning 3D Display
Hiroyo Ishikawa; Hideo Saito
Proceedings of 18th International Conference on Artificial Reality and Telexistence (ICAT2008), Nov. 2008, [Reviewed] - Point Cloud Representation of 3D Shape for Laser-Plasma Scanning 3D Display
Hiroyo Ishikawa; Hideo Saito
IECON 2008: 34TH ANNUAL CONFERENCE OF THE IEEE INDUSTRIAL ELECTRONICS SOCIETY, VOLS 1-5, PROCEEDINGS, Nov. 2008, [Reviewed] - Relation between psychological amount in polyhedron recognition and 3D object image entropy
Hiroyo Ishikawa; Yukio Sato
Systems and Computers in Japan, Mar. 2005, [Invited] - Relation between Psychological Amount in Polyhedron Recognition and 3-D Object Image Entropy.
石川尋代; 佐藤幸男
電子情報通信学会論文誌 D-2, Jan. 2003, [Reviewed]
In this paper, the relationship between the psychological information quantity quantified by humans regarding the perceptual ease of object shapes that vary with viewing direction and the object image entropy representing physical information quantity is examined. Here, we propose constructing the psychological information quantity representing the psychological measure of how easy an object shape is perceived to be using the paired comparison method and Thurstone's comparative judgment method. Object image entropy, on the other hand, is a one-dimensional measure defined based on the number of visible faces within an object image and their area ratios. It represents an information quantity reflecting both the qualitative aspect of the object image and the quantitative characteristic of area ratios. Here, we limit the objects to polyhedra, measure the psychological information quantity and object image entropy for various types of shapes, and examine the relationship between the two. As a result, although there are some exceptions, a monotonically increasing relationship was observed between the psychological information quantity and the object image entropy for many types of objects.
MISC
- [掲載予定] MoSaICによる写真を用いた三田演説館の関係記述とそのビジュアライゼーションに関する試み
Mar. 2026
Lead - デジタルコンテンツ間の多様な関係を可視化するMoSaIC(Museum of Shared and Interactive Cataloguing)の研究開発
May 2024 - Organizing Digital Content Information in the MoSaIC System
Hiroyo Ishikawa
DMC Review, Mar. 2019
Lead, Symposium Record: “8th Symposium on the Cultural Dissemination and Deepening of Digital Knowledge — Rethinking Metadata” A presentation on conceptual approaches to metadata in the organization (i.e., relational structuring) of digital content within the MoSaIC system., Research Institute for Digital Media and Content, Keio University - Modeling Relationships in MoSaIC: An Engineer’s Perspective
Hiroyo Ishikawa
慶應義塾大学DMC紀要, Mar. 2018
Lead, Symposium Presentation Record: “7th Symposium on the Cultural Dissemination and Deepening of Digital Knowledge — Toward Contextual Networking in Decentralized Museums” A lecture tracing five years of research and development on the MoSaIC project, centered on the concept of “relationships” and their modeling. - Paneru disukasshon : DMC ga egaku korekara no myujiamu sekai
松田 隆美; 安藤 広道; 石川 尋代; 生貝 直人; 金子 晋丈
慶應義塾大学DMC紀要, Mar. 2016
Panel discussion held at the symposium “The 5th Cultural Dissemination and Deepening of Digital Knowledge: Expanding New Museum Worlds through Multifaceted Archives.” This record summarizes the discussion on how to construct digital museums using the visualization system for multifaceted archives developed by DMC., 慶應義塾大学デジタルメディア・コンテンツ統合研究センター - Development of the MoSaIC System: Visualization of Digital Content Space
Hiroyo Ishikawa
慶應義塾大学DMC紀要, Mar. 2015
This paper describes the development of the MoSaIC browsing system, which interactively visualizes diverse relationships among digital contents and uses the results as a visual interface for content selection. From the developer’s perspective, it explains specific issues encountered over the three years of conceptualization and development, along with the solutions, 慶應義塾大学デジタルメディア・コンテンツ統合研究センター - Paneru disukasshon : MoSaIC ni yoru tamenteki akaivu eno chosen
嘉村 哲郎; 池田 真弓; 石川 尋代; 都倉 武之; 金子 晋丈
慶應義塾大学DMC紀要, Mar. 2015
At the symposium “The 4th Challenge Toward Multifaceted Archives through MoSaIC: Toward Cultural Dissemination and Deepening of Digital Knowledge,” a panel discussion unfolded around the structure of information and the diversity of interpretation. Using examples from war archives, the multifaceted nature of materials was highlighted, and the conceptual underpinnings of nodes, groupings, and seamlessness were examined. MoSaIC was presented as a potential site of knowledge that inherently embraces cultural inquiry., 慶應義塾大学デジタルメディア・コンテンツ統合研究センター - MoSaIC-II: Tayona kankei o kashikasuru bijuaru intafesu o mochiita dejitaru kontentsu etsuran shisutemu : dejitaru kontentsu no tayona kankei no moderuka to sono kashika no kokoromi
石川 尋代
慶應義塾大学DMC紀要, Mar. 2014
We developed MoSaIC-II, a browsing system that interactively visualizes diverse relationships among digital contents and uses the results as a visual interface for content selection. We created MoSaIC catalogues and conducted browsing experiments to confirm the system’s effectiveness., 慶應義塾大学デジタルメディア・コンテンツ統合研究センター
Books and other publications
Lectures, oral presentations, etc.
- MoSaIC: Connecting digital contents by various contexts (様々なコンテクストによってデジタルコンテンツを繋ぐ)
Jul. 2018 - 様々な記録を繋げる多面的アーカイヴ-MoSaICの試み
Nov. 2015, [Invited] - Design and Prototype of Museum of Shared and Interactive Cataloguing
Hiroyo Ishikawa; Kunitake Kaneko
NODEM(Nordic Digital Excellence in Museums)2014 Conference & Expo http://repo.nodem.org/?objectId=360, Dec. 2014
To enable users to view digital content from various perspectives and navigate its context, we designed the Museum of Shared and Interactive Cataloging (MoSaIC) system. This system consists of a decentralized global catalog sharing mechanism and a catalog visualization mechanism, allowing users to easily compare and understand differences in viewpoints and contexts. In addition to a presentation, the actual system was exhibited for visitors to experience. - A digital content viewing system using polymorphic visualization of relations between content as a visual interface
ISHIKAWA Hiroyo; MIYASHITA Yamato; KANEKO Kunitake; SAITO Hideo; MATSUDA Takami
ITE Technical Report, Mar. 2014
In this paper we propose a digital content viewing system which visualizes the polymorphic relationships between digital content interactively and uses it as a visual interface. The relationships are expressed by classifying and corresponding of the content and shown by directed graphs. On the graphs, nodes mean the content files and the groups, and edges mean group members or correspondences. The graphs are archived as relation catalogues separately from content files. We visualize the relations of content on a directed graph. To visualize the graphs, we use a method of locating nodes by horizontal directions and height. Nodes linked with a target node will be located on a horizontal circle and their height is decided by their relation catalogue. After deciding the node location, nodes and links are drawn using 3D graphics and animation. The visualization result called "polymorphic view" is used as a visual interface for selecting content. Selecting content on the visualized graph will expand new links and new content nodes additionally. We can show another content by tracing graph links. It becomes easy to understand the relations between digital content, because we can change the direction and size of the directed graph. - Interactive display of shining gilded illumination in Keio Gutenberg bible
Hiroyo Ishikawa; Fumihiko Kamemura; Kunitake Kaneko; Hideo Saito; Takami Matsuda
JADH(Japanese Association for Digital Humanities) 2012 Conference “Inheriting Humanities", Sep. 2012
This study proposes an interactive display system for observing rare books with gilded illuminations by virtually changing light direction. A 3D model of the Gutenberg Bible is displayed on iPad with touch-based control. - Surface representation of a polygonal model for aerial 3D display
ISHIKAWA Hiroyo; SAITO Hideo
ITE Technical Report, Jul. 2010
In this paper, a method of creating contents data for a new spatial 3D display which can draw dots of light to represent an object in space is proposed. The Aerial 3D display that we use in this research can create a dot of light at 50kHz, and can display it in an arbitrary position by vector scanning. Since it can display a higher density of dots of light, we expect that the desktop system can represent more complex objects, such as 3D surface models consisting of polygonal patches, which are generally used in computer graphics. The proposed method represents the surface with cross sections of an object against the vertical direction and generates a point sequence data for vector scanning. From the experiments of drawing, it can be seen that polygonal models can successfully be drawn by the proposed method. - Representation of Characters for Laser-Plasma Scanning 3D Display
ISHIKAWA Hiroyo; SAITO Hideo
IEICE technical report, 06 Mar. 2009
A new device which can generate point sequence of plasma luminous bodies has been developed for a new type 3D display. In this system, the plasma luminous bodies sequentially appear and disappear, 1000 times a second. Therefore there is a restriction on the number of the plasmas which human can perceive simultaneously. In this research, we represent characters variously by changing the number of the plasmas and the drawing order and display them on our system. The easiness of recognizing characters is measured in the experiments for human. It can be seen that the way of drawing several times by reducing the density of points is stable to recognize a character, even though the drawing time is the same. Furthermore, the character string is considered, and it can be seen that the way of displaying several characters is an appropriate method. A method of representing characters is proposed in consideration of the result. - Point Cloud Generation for Displaying Face Using Free Space Display by Laser Plasma
AOKI Satoshi; NOZICK Vincent; ISHIKAWA Hiroyo; SAITO Hideo
IPSJ SIG Notes. CVIM, 01 May 2008
Recent research on laser plasma has made possible to draw light dots on the air. This new technology, called "Free Space Display by Laser Plasma" , enables to draw 3 dimensional monochrome images on the air using plasma dots. This paper presents a method to generate a point cloud from an image and its corresponding depth map. Our method especially focuses on human faces rendering. Due to the laser device limitation on the number of points to be drawn, these points should be chosen carefully, such they can reliably depict the human face. During a pre-processing stage on the input videos, a depth-map of the face is computed on every frame using a stereo method. We also perform a background subtraction on the input images associated to the depth-maps. Then, our method selects significant points to visually represent the face using a weighted method applied on points candidate provided by an edge detection pass. Then the selected points are combined with the depth map to generate 3D points that can be send to the laser display. - GPU for Image Processing and Computer Vision
NOZICK Vincent; ISHIKAWA Hiroyo; DE SORBIER Francois
IPSJ SIG Notes. CVIM, 01 May 2008
Contrary to CPU, GPU capability continues to increase and reaches unexpected performances for image processing tasks. Indeed, GPU is mainly designed for computer graphics applications and can deal with 2D texture or 3D operations very efficiently. These abilities are well suited to speed up real-time image processing and computer vision applications. Contrary to the first generations of shaders, the latest shaders languages become more and more easy to use. However, some basic knowledge about shaders is required for an efficient use of the GPU. This paper presents an overview of how to use shaders for image processing and computer vision. The first part details the computer graphics rendering pipeline and shaders generalities. This part also overview the existing shader languages and especially details GLSL. These explanations include shaders programming but also the shader loading and compilation. This part aims to underline that shaders require a specific program design. Then, this paper presents some practical applications related to image processing and computer vision. These sections especially deal with colour manipulation, geometric applications but also with virtual reality and general purpose GPU methods. Finally, this document provides a part on how to start with shaders, including technical and theoretical recommendations. - Efficient Expression of 2D shapes using points in consideration of human visual characteristics
ISHIKAWA Hiroyo; SAITO Hideo
IEICE technical report, 10 Jan. 2008
A new device has been developed for a three-dimensional aerial display that generates plasma light in mid-air to render point clouds of light. This study proposes a method for efficiently representing 2D contours using point clouds, based on human visual characteristics. In particular, regions with high curvature are assigned higher point density to enhance shape recognizability. - 3D Human Body Measurement and Modeling
ISHIKAWA Hiroyo; SATO Yukio
Technical report of IEICE. Multimedia and virtual environment, 20 Jan. 2005, The Institute of Electronics, Information and Communication Engineers
We propose a method for constructing the surface patch model for human body. This model is built from 3D shape data that is measured using 3D human body measurement system. The system is built combining the small range finder that can measure object shape at high speed. Here, as an example, we construct a model about a head has a lot of personal information. And a natural motion is produced by adding a motion parameter, about speaking face, to the generated model. - 不完全な距離データからの足モデル構築
Jul. 2004 - Relation between Psychological Amount in Polyhedron Recognition and 3-D Object Image Entropy
Hiroyo Ishikawa; Yukio Sato
Sep. 2003
This study evaluates polyhedral recognizability by comparing object image entropy—defined as the amount of visual information—with perceptual psychological measures. Object image entropy is calculated based on the number and area ratio of surface regions, while perceptual measures are quantified using Thurstone’s paired comparison method. Experimental results revealed a positive correlation between entropy and recognizability, suggesting that surface information plays a key role in polyhedral perception. - Inference for the optimal view direction in polyhedron recognition
ISHIKAWA Hiroyo; SATO Yukio
Proceedings of the IEICE General Conference, 03 Mar. 2003, The Institute of Electronics, Information and Communication Engineers
This study conducted a simulation to explore the viewpoint from which a polyhedron’s shape is most easily recognized. By observing a 3D-modeled polyhedron from a given direction, object image entropy was calculated. The system then automatically searched nearby viewpoints for higher entropy values, updating the viewpoint accordingly. This process was repeated until the viewpoint stabilized, resulting in the most recognizable view of the polyhedron. - Relation between psychological distance and physical feature amount of a plane shape in the polyhedron
ISHIKAWA Hiroyo; OKUMURA Kouji; SATO Yukio
IPSJ SIG Notes. CVIM, 08 Nov. 2001, Information Processing Society of Japan (IPSJ)
This study focuses on surface similarity as one of the factors for constructing a scale of polyhedral recognizability, evaluated through human visual processing characteristics. Using a triadic comparison method and multidimensional scaling, psychological distances between 2D surface shapes were derived and compared with physical distances calculated from various shape features. The study explores which physical attributes influence perceived similarity. - Influence of Dummy Binocular Disparities and Monocular Clues on the Depth Perception
ARAKAWA Naofumi; ISHIKAWA Hiroyo; AMANO Toshiyuki; SATO Yukio
Technical report of IEICE. PRMU, 08 Nov. 2001, The Institute of Electronics, Information and Communication Engineers
The phenomenon in which depth feeling increases is known for many years by giving dummy binocular disparities to a single picture, and getting rid of the vergence. Although the reason is not solved yet, if solved, it is clear to take a big step toward realization of stereoscopic vision from a single picture. Then, towards realization of stereoscopic vision by the single picture, the influence on the depth perception which dummy binocular disparities and monocular clues do was investigated quantitatively. From the result, depending on the intensity of monocular clues, it became clear to enable depth perception of the same grade as the case where a stereo picture is used also by use of the picture which added dummy binocular disparities. - Relation between psychological scale in polyhedron recognition and 3-D object image entropy
ISHIKAWA Hiroyo; SATO Yukio
Technical report of IEICE. PRMU, 10 Nov. 2000, The Institute of Electronics, Information and Communication Engineers
In 3-D object recognition, we can select a view direction that is easy to recognize quickly. Therefore there is a psychological factor about easiness to perceive object shape, and we select a view direction using the psychological scale. We defined the object image entropy. It is the amount of information that image includes. This entropy is a scale calculated from number of surface observed in an image and their area ratio. In this paper, we measure about easiness of perception that changes by view direction and consider that relation between entropy and psychological scale.
Affiliated academic society
Research Themes
■University education and qualification information
