TY - GEN
T1 - Recognition of human emotion using sensor agent robot for interactive and adaptive living spaces
AU - Murata, Sozo
AU - Mita, Akira
PY - 2011/5/26
Y1 - 2011/5/26
N2 - Safer, more comfortable and energy-efficient living spaces are always demanded. However, most buildings are designed based on prescribed scenarios so that they do not act on abrupt changes of environments. We propose "Biofication of Living Spaces" that has functions of learning occupants' lifestyles and taking actions based on collected information. By doing so, we can incorporate the high adaptability to the building. Our goal is to make living spaces more "comfortable". However, human beings have emotion that implies the meaning of "comfortable" depends on each individual. Therefore our study focuses on recognition of human emotion. We suggest using robots as sensor agents. By using robots equipped with various sensors, they can interact with occupants and environment. We use a sensor agent robot called "e-bio". In this research, we construct a human tracking system and identified emotions of residents using their walking information. We focus on the influences of illuminance and sound. We classified emotions by calculating the distance of the mapped points in comfortable and uncomfortable spaces with parametric eigen space method, in which parameters are determined by a mapping of tracks in the space. As a method of pattern recognition, a weighted k-nearest neighbor is used. Experiments considering illuminance and sound environments, illustrates good correlation between emotion and environments.
AB - Safer, more comfortable and energy-efficient living spaces are always demanded. However, most buildings are designed based on prescribed scenarios so that they do not act on abrupt changes of environments. We propose "Biofication of Living Spaces" that has functions of learning occupants' lifestyles and taking actions based on collected information. By doing so, we can incorporate the high adaptability to the building. Our goal is to make living spaces more "comfortable". However, human beings have emotion that implies the meaning of "comfortable" depends on each individual. Therefore our study focuses on recognition of human emotion. We suggest using robots as sensor agents. By using robots equipped with various sensors, they can interact with occupants and environment. We use a sensor agent robot called "e-bio". In this research, we construct a human tracking system and identified emotions of residents using their walking information. We focus on the influences of illuminance and sound. We classified emotions by calculating the distance of the mapped points in comfortable and uncomfortable spaces with parametric eigen space method, in which parameters are determined by a mapping of tracks in the space. As a method of pattern recognition, a weighted k-nearest neighbor is used. Experiments considering illuminance and sound environments, illustrates good correlation between emotion and environments.
KW - Biofication of Living Spaces
KW - human emotion
KW - parametric eigen space method
KW - sensor agent robot
UR - http://www.scopus.com/inward/record.url?scp=79956303806&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=79956303806&partnerID=8YFLogxK
U2 - 10.1117/12.880246
DO - 10.1117/12.880246
M3 - Conference contribution
AN - SCOPUS:79956303806
SN - 9780819485434
T3 - Proceedings of SPIE - The International Society for Optical Engineering
BT - Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2011
T2 - Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2011
Y2 - 7 March 2011 through 10 March 2011
ER -