2P1-M10 Generation of human behavior models is based on the transition relations of between the HMMs and statistical processes
Kae Doki, Takahiro Hirai, Takashi Hattori, Shinji Doki, Yuki Funabora, Akihiro Torii, Suguru Mototani · The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec) · 2015
We are aiming to automatic generation of high human behavior model reproducible from huge data about human behavior. There-fore, it is necessary to extract a part of the basic elements of a series of time-series data. In this paper, a series of time-series data is quantized in the HMMs, and to network the transition between high reproducibility HMMs. Extracts the basic elements based on the network to generate the behavior model.