業績

2018 Fiscal Year

Books and Proceedings

Issei Fujishiro, Jinwook Seo (editors)
Proceedings of PacificVAST 2018
The Journal of Visual Informatics, Vol. 2, No. 1, pp. 1-110, March 2018
These are the proceedings of PacificVAST 2018 Workshop, which was held at Kobe University on April 10th, 2018. Prof. Fujishiro served as a workshop chair and a guest editor in chief for the special issue. These proceedings start with the guest editors’ foreword, followed by nine papers accepted by the workshop.PacificVAST has been held as a workshop of IEEE PacificVis Symposium since 2013. The past five year’s programs consisted only of invited talks, but PacificVAST 2018 is the first year that called for original papers and published its proceedings as the special issue of Elsevier Journal of Visual Informatics.

Academic Papers in English

Yusuke Ishikawa, Issei Fujishiro
TideGrapher: Visual analytics of tactical situations for rugby matches
The Journal of Visual Informatics, Vol. 2, No. 1, pp. 60-70, March 2018 [DOI: 10.1016/j.visinf.2018.04.007]
Various attempts at exploiting information visualization for sports have recently been reported in the literature, although it is still challenging to analyze continuous ball matches. In this paper, we propose a novel visual analytics system, called TideGrapher, to track the transition of tactile situations in a rugby match. With a particular focus on the side position of the ball, we designed a dedicated spatial substrate based on the spatio-temporal trajectory of the ball and provided a set of basic interactions. Quantitative analysis was strengthened by adding a new index, called initiative, to commonly used possession (ball occupation) and territory (dominance of territory). The feasibility of the proposed visual analytics system was proven empirically through application to datasets from real amateur and professional matches.

Yuriko Takakura, Masanori Nakayama, Issei Fujishiro
Visual Analysis for the Compositional Process of Composers in Spectral School
IIEEJ Transactions on Image Electronics and Visual Computing, Vol. 6, No. 1, pp. 22-30, June 2018
We propose a visual analysis approach for exploring the compositional process of composers in the Spectral School, with a particular focus on its sub-processes of sound analysis and synthesis. Spectral music has been one of the significant trends in contemporary music since the 1970s. Composers in the Spectral School use the acoustic properties of sounds as the basis of their compositional materials. One of the representative software systems they use is AudioSculpt, developed by IRCAM. We develop an accompanying system with AudioSculpt to manage the history of creating sounds with spectral analysis of sound materials. The salient feature of our system lies in its tessellated spectrogram space whose axes represent the elapsed time of the sound and the progress of composition. On the pixel-oriented, spatial substrate, the system makes it possible for users to analyze the compositional processes by using dedicated interactive manipulations. The present system can also be regarded as an initial attempt at managing the provenance of time-series events in the music visualization field. Our approach is intended to pave the way for composers and musicologists to analyze, develop, and share the compositional methodologies.

Issei Fujishiro, Naoko Sawada, Masanori Nakayama, Hsiang-Yun Wu, Kazuho Watanabe, Shigeo Takahashi, Makoto Uemura
TimeTubes: Visual Exploration of Observed Blazar Datasets
Journal of Physics: Conference Series, Vol. 1036, No. 1, Article No. 012011, June 2018 [DOI: 10.1088/1742-6596/1036/1/012011]
Blazars are attractive objects for astronomers to observe in order to burrow into the magnetic field in the relativistic jet. This paper presents TimeTubes as a novel visualization scheme that allows astronomers to interactively explore characteristic temporal variation patterns in observed blazar datasets. In the TimeTubes spatialization, the two Stokes parameters and their errors with a common timestamp are transformed into an ellipse. A series of such ellipses are aligned in parallel along the timeline to form a 3D volumetric tube. The resulting tube is then colorized by the observed intensities and colors of the blazar, and finally volume-rendered. A designated user interface is provided with visual exploration functions according to Shneiderman’s Visual Information Seeking Mantra. In the latest version, an auxiliary mechanism, called visual data fusion, was incorporated to ameliorate data- and mapping-inherent uncertainties for more efficient and effective visual exploration.

International Conference Proceedings

Naoko Sawada, Masanori Nakayama, Makoto Uemura, Issei Fujishiro
TimeTubes: Automatic Extraction of Observable Blazar Features from Long-Term, Multi-Dimensional Datasets
in Proceedings of IEEE Scientific Visualization Conference (SciVis), October 2018
Blazars are attractive objects for astronomers to observe in order to demystify the relativistic jet. Astronomers need to classify characteristic temporal variation patterns and correlations of multi-dimensional time-dependent observed blazar datasets. Our visualization scheme, called TimeTubes, allows them to easily explore and analyze such datasets geometrically as a 3D volumetric tube. Even with TimeTubes, however, data analysis over such long-term datasets costs them so much labor and may cause a biased analysis. This paper, therefore, attempts to incorporate into the current prototype of TimeTubes, a new functionality: feature extraction, which supports astronomers’ efficient data analysis by automatically extracting characteristic spatiotemporal subspaces.

International Conference Posters (Peer-Reviewed)

Kazuki Miyazaki, Issei Fujishiro
Automatic Generation of Artworks Using Virtual Photoelastic Material
in Proceedings of ACM SIGGRAPH 2018 Posters, Vancouver (Canada), August 2018 [DOI: 10.1145/3230744.3230797]
Photoelasticity is known as one of the phenomena related to polarization and is defined as the change in birefringence of transparent material when internal force is applied. Interference fringes appear by irradiating the material with polarized light when viewing it through the polarizer. In this study, we attempt to apply the concept of photoelasticity to generative art. Assuming there is virtual stress distribution in the two-dimensional material, our method automatically generates artworks with photoelasticity. A GPU-based acceleration of the current implementation is also discussed.