Pregled bibliografske jedinice broj: 890660
Brain Computer Interface Communicator : A Response to Auditory Stimuli Experiment
Brain Computer Interface Communicator : A Response to Auditory Stimuli Experiment // MIPRO 2017, 40th International Convention Proceedings / Biljanović, Petar (ur.).
Rijeka: Hrvatska udruga za informacijsku i komunikacijsku tehnologiju, elektroniku i mikroelektroniku - MIPRO, 2017. str. 534-537 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 890660 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Brain Computer Interface Communicator : A Response to Auditory Stimuli Experiment
Autori
Jadav, Guruprasad Madhale ; Batistić, Luka ; Vlahinić, Saša ; Vrankić, Miroslav
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
MIPRO 2017, 40th International Convention Proceedings
/ Biljanović, Petar - Rijeka : Hrvatska udruga za informacijsku i komunikacijsku tehnologiju, elektroniku i mikroelektroniku - MIPRO, 2017, 534-537
ISBN
978-953-233-093-9
Skup
MIPRO 2017 40th Jubilee International Convention
Mjesto i datum
Opatija, Hrvatska, 22.05.2017. - 26.05.2017
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
BCI, communicator, Locked-in, Auditory stimuli
Sažetak
This paper describes a communication system, designed to help people with disabilities to communicate with others. In this case, it is built on Event Related Potentials (ERP) which occur as a response to stimuli. The stimuli are auditory answers to the asked questions, where one of the stimulus is the correct answer. The stimulus which is the possible correct answer is the target stimulus and other stimuli are considered as non- targets. The only reliable information available is the onset of the stimuli, based on this information the spatio-temporal pattern for the target and non-targets is expected to be uncorrelated to the ongoing brain activity and with each other. This allows us to train the spatial filter coefficients for the know targets, so that it can use this trained data set to correctly classify targets. Results are obtained by optimizing the spatial filter parameters prior to classification, and the classification results are analyzed for several healthy participants.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika
POVEZANOST RADA
Ustanove:
Tehnički fakultet, Rijeka
Citiraj ovu publikaciju:
Časopis indeksira:
- Scopus