초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

EEG based brain-computer interface directly transfers information of the brain signal into a computer or machine through electrical pathway without limb movement. In this field, common spatial pattern (CSP) algorithm is a very well-known and efficient method extracting discriminative features from t...

EEG based brain-computer interface directly transfers information of the brain signal into a computer or machine through electrical pathway without limb movement. In this field, common spatial pattern (CSP) algorithm is a very well-known and efficient method extracting discriminative features from two different conditioned brain signals. Since CSP was firstly proposed in BCI about a decade ago, many variants of CSP algorithms have been developed. Among them, common spatio-spectral pattern (CSSP) and invariant common spatial pattern (iCSP) have been reported to show relatively better performance than conventional CSP because they can consider more information and more efficient noise suppression than CSP. We compared these kinds of CSP algorithms in terms of classification accuracy, and suggested new invariant CSSP which takes both of advantages from CSSP and iCSP by adding noise suppression term taking into account non-stationarity to CSSP. Comparative study with 48 subjects showed that our new algorithm outperformed CSSP.

주제어