Within Ritsumeikan University’s Art Research Center, Professor Hachimura Kozaburo of the College of Information Science and Engineering is pioneering research in the archiving of Japanese classical dance and performing arts. An expert in image processing and computer graphics, who has also had significant experience in the humanities through four years of work at the National Museum of Ethnology in Osaka, Professor Hachimura was instrumental in the establishment of the Art Research Center in 1998 and currently serves as the sub-leader of the Kyoto Art and Entertainment Innovation Research COE Program, which is based in the center.
Professor Hachimura’s research objective is the digital archiving of intangible cultural assets, such as the classical Japanese dance and theatrical arts that originated in Kyoto. Since receiving a government grant to purchase an optical motion capture system in 2000, his research team has been recording and archiving dance movements of the human body and kabuki and noh performances. Motion capturing allows researchers to acquire time-series data of three-dimensional coordinates for various body parts. Until recently, this technology has been mainly used in the commercial field for making computer games and films and in the academic field for biomechanics and sports sciences research. Using motion capturing technology for the purpose of archiving intangible cultural property is a relatively new idea.
What sets the Art Research Center’s digital archiving approach apart from other universities in Japan is its location in Kyoto, which provides access to the highest quality of traditional Japanese performers in the world. The Art Research Center works closely with the Katayama House of the Kanze School of Noh and the Inoue School of Kyoto Classical Dance?which are the premier institutions in each field?as well as with several top kabuki actors. Thus, the data recorded and archived is of the highest standard. The research team has already captured noh performances, classical Japanese dance, and, for comparative research, modern dance, ballet, and classical dances from various regions of the world. Utilizing this quantitative data to produce computer graphic animation and multimedia educational materials is an important application of the research.
Professor Hachimura’s main role in this project is the development of a method for analyzing human body motion data. One project he is working on is measuring the differences in dance performances between experts and beginners and developing a better way to teach dance through virtual reality instruction. He is also particularly interested in comparisons between human body motion across different cultures and genres of dance.
The other main issue in his research is that of data retrieval. While catalogued data, such as the name of the dance or the dancer, is certainly one means of retrieving data, Professor Hachimura wants to also incorporate a system to retrieve data based on the similarity of a movement itself. In order to conduct “similarity retrieval,” it is necessary to determine the degree of similarity among body motion data. Differences in speed, body type, and size must also be taken into account. Professor Hachimura has developed a prototype system for similarity retrieval using a dynamic programming (DP) matching method?a common technique in voice recognition technology that is in this case applied to motion. This system conducts motion retrieval based on similarities between sets of body motion data.
Labanotation, a system that uses graphic symbols to record human body motion in dance (comparable to a music score), also plays a part in data retrieval. Labanotation scores can be automatically generated from data obtained during motion capturing and can be used as a method to identify various dance patterns. While Labanotation is not as accurate as motion capturing, which can measure motion down to one tenth of a millimeter, both methods are useful and will be combined in the database.
The complete digital archive will be accessible online once it has been completed. Professor Hachimura and his colleagues plan to continue to explore the traditional Japanese performing arts by using the archive to conduct both quantitative and qualitative research into the aesthetics of dance.
Within Ritsumeikan University’s Art Research Center, Professor Hachimura Kozaburo of the College of Information Science and Engineering is pioneering research in the archiving of Japanese classical dance and performing arts. An expert in image processing and computer graphics, who has also had significant experience in the humanities through four years of work at the National Museum of Ethnology in Osaka, Professor Hachimura was instrumental in the establishment of the Art Research Center in 1998 and currently serves as the sub-leader of the Kyoto Art and Entertainment Innovation Research COE Program, which is based in the center.
The other main issue in his research is that of data retrieval. While catalogued data, such as the name of the dance or the dancer, is certainly one means of retrieving data, Professor Hachimura wants to also incorporate a system to retrieve data based on the similarity of a movement itself. In order to conduct “similarity retrieval,” it is necessary to determine the degree of similarity among body motion data. Differences in speed, body type, and size must also be taken into account. Professor Hachimura has developed a prototype system for similarity retrieval using a dynamic programming (DP) matching method?a common technique in voice recognition technology that is in this case applied to motion. This system conducts motion retrieval based on similarities between sets of body motion data.
