10채널 뇌파를 이용한 감성 평가에 관한 연구

A Study on the Human Sensibility Evaluation Using 10-channel EEG

  • 강동기 (청주대학교 전자정보통신반도체공학부) ;
  • 김흥환 (청주대학교 전자정보통신반도체공학부) ;
  • 김동준 (청주대학교 전자정보통신반도체공학부) ;
  • 고한우 (한국표준과학연구원)
  • Kang, Dong-Kee (School of Electronic, Semiconductor, Computer & Communication Eng., Chongju Univ.) ;
  • Kim, Heung-Hwan (School of Electronic, Semiconductor, Computer & Communication Eng., Chongju Univ.) ;
  • Kim, Dong-Jun (School of Electronic, Semiconductor, Computer & Communication Eng., Chongju Univ.) ;
  • Ko, Han-Woo (Korea Research Institute of Standards and Sciences)
  • 발행 : 2001.11.24

초록

This paper describes a method of human sensibility evaluation for pleasant and unpleasant environments. Conditions of the environment are room temperature and humidity. Changing the conditions, 10-channel EEG signals for 4 subjects are collected. Linear predictor coefficients of the recorded EEGs are extracted as the feature parameter of human sensibility. A neural network-based human sensibility estimation algorithm is developed. The developed algorithm showed good performance in the pleasantness evaluation. The neural network output produced accurate states of pleasantness sensibility. Subject-independent test showed similar results with subject-dependent test.

키워드