May 10, 2011
The 104th GCOE Seminar
1. “Revealing Visual Hypotheses from Japanese Historical Diary”
Lecturer: Alejandro Toledo Nolasco (RA/Digital Archive Technology Research Group/ Ritsumeikan University Graduate School of Science and Engineering D3)
2. “Similarity Retrieval of Motion Capture Data Based on Derivative Feature”
Lecturer: Worawat Choensawat (RA/Digital Archive Technology Research Group/Ritsumeikan University Graduate School of Science and Engineering D3)
3. “Interpersonal Timing of Up-Down Dance Movements: Focused on shoulders and knees movements”
Lecturer: Shikanai Naho (RA/Japan Society for the Promotion of Science, Special Researcher/Digital Archive Technology Research Group/Ritsumeikan University Graduate School of Science and Engineering D2)
Time: May 10 (Tues.) 18:00-19:30 (6th Period)
Place: (Kinugasa) Ritsumeikan University Art Research Center Multipurpose Room (BKC) Please use the Internet (Power Live)
Free of charge (no reservation required)
* The general public may also participate in the seminar through the Internet.
* Those who wish to participate on the Internet, please contact tkiri@fc.ritsumei.ac.jp (Attn: Kirimura)
* GCOE research members and those who already have ID and passwords do not need to contact us.
* Starting in the afternoon on the day of the event, materials can be viewed at the following link (for a limited time):
http://www.arc.ritsumei.ac.jp/dhjac/GCOESeminar/haihusiryo-index.html
To view presentation abstracts, please click on “more” below.
1. Alejandro Toledo Nolasco, “Revealing Visual Hypotheses from Japanese Historical Diary”
We present an interactive visualization system which allows users to explore a collection of personal names extracted from the Hyohanki diary (Taira no Nobunori, late Heian era, 1132-1184). In our current research on stacked graphs, our system has been equipped with a novel recommender engine in order to reveal what we call visual hypotheses, particular subsets in the Hyohanki-name space that potentially represent indicators of name occurrence trends. Our system reveals visual hypothesis using a method based on dwell time as implicit interest indicator. We believe that the revealed visual hypotheses might serve users’ needs in their analytical tasks.
2. Worawat Choensawat, “Similarity Retrieval of Motion Capture Data Based on Derivative Feature”
This paper, we propose a similarity retrieval of motion capture data in which a new feature extraction is introduced for the improvement of similarity search. For similarity search, joint speed has been mainly used as features of a particular motion. We use not only the magnitude of speed but also a pattern of speed change. We find the pattern of changing joint speed in a short period of time with the derivative of joint speed. In our experiments, we found that our proposed feature extraction could improve the search precision and time. The average precision was greater than 90% and its computational time of 17 seconds on a dataset of 225 motion clips with total of 81,851 frames from CMU's database.
3. Shikanai Naho, “Interpersonal Timing of Up-Down Dance Movements: Focused on shoulders and knees movements”
In order to clarify the effect of interactions with others in dance, we carried out a motion- capture experiment with two dancers. We set the condition of mutually looking at each other face-to-face as a synchronized action. In turn, we set the condition of not being face-to-face, where only one person could view the other, as an unsynchronized action. In this research, we analyzed those changes and timings, focusing on the movements of shoulders and knees. As a result, we confirmed that whether face-to-face or not, synchronous actions were fewer than unsynchronized actions in regards to the timing of the two dancers, and displacement was small.
1. “Revealing Visual Hypotheses from Japanese Historical Diary”
Lecturer: Alejandro Toledo Nolasco (RA/Digital Archive Technology Research Group/ Ritsumeikan University Graduate School of Science and Engineering D3)
* Video distribution stopped due to the wishes of the speaker. (2011.5.11)
3. “Interpersonal Timing of Up-Down Dance Movements: Focused on shoulders and knees movements”
Lecturer: Shikanai Naho (RA/Japan Society for the Promotion of Science, Special Researcher/Digital Archive Technology Research Group/Ritsumeikan University Graduate School of Science and Engineering D2)
