OUR CURRENT PROJECTS
DIGITALISATION OF TRADITIONAL DANCES USING MOTION CAPTURE
DJOGET is our pilot project in documenting and digitising traditional dances of Indonesia. The project is started in 2016 and currently have documented several Javanese traditional dances, in collaboration with Institut Seni Indonesia Yogyakarta. Our first step is documenting the dances using recording media (video and audio), and in further, the movement of each dance will be stored as 3D motion data using motion capture devices.
Our goal is preserving the details of a dance, not just through instructive means, but enables a visual recreation of the true movements, thus preserving the originality of dance details which will be timeless, learnable by new generation of dancers, without being dependent to the living media (maestros) to teach the dance.
The project comes out as a result of the more and more less regeneration in Indonesian traditional dancers, the less interests of Indonesian younger generations on their ancestry roots, which eventually may cause the knowledge of traditional dances thoroughly eliminated. By preserving the detailed movements in video and 3D motion form, we are hoping to create a repository, safely stored, learnable and researchable for the future generations.
PREPROCESSING OF JAVANESE MANUSCRIPTS
A Collaborative Study with Project TRAWACA
In the same spirit as DJOGET, we are collaborating with Project TRAWACA from Universitas Kristen Duta Wacana, which in search on optimal OCR method for transliterating old Javanese manuscript. Project TRAWACA itself will be a lifelong project, continuously studying the best efforts in digitising the contents of the manuscripts, from the perspective of information technology, including using Natural Language Processing for post-processing.
Currently, the project have collected and digitised about a hundred manuscript pages from various sources (museum, private collection, flea markets) and at the moment is working for the preprocessing of digital images as a preparation for the OCR process. Preprocessing studied since 2016 is degradation reduction, background segmentation and Javanese corpus building. This year we are working on Javanese syllabification and character segmentation.
Our dream by this project is creating a transliteration engine liable in acquisiting the content from a scanned manuscript and store it digitally, whether in digital Javanese script or its Latin transilteration. We are going beyond just image for digitalisation.