Work

Wearable Forehead EOG System and Comfortable Auditory Stimulation Method for Human Computer Interface

Year: 2018
Type: dissertation
Abstract: the sounds of water streaming and cicadas singing were used in the natural sound sessions. Six healthy subjects participated in the experiment. Electroencephalograms were recorded using four electrodes (Cz, Oz, T7 and T8). Seven sessions were performed using different stimuli. The spectral power at 38 and 42 Hz and their ratio for each electrode were extracted as features. Linear discriminant analysis was utilized to classify the selections for each subject. In an offline analysis, the average classification accuracy with a modulation index of 1.0 were 89.67% and 87.67% using music and natural sounds, respectively. In online experiments, the average classification accuracies were 88.3% and 80.0% using music and natural sounds, respectively. Using the proposed method, we obtained significantly higher user-acceptance scores, while maintaining a high average classification accuracy. \nTo date, the human–computer interface has achieved many technological advances. Henceforth, it will also be necessary to improve user convenience in consideration of ergonomic factors' increased performance. It is expected that the proposed system will be able to help disabled people accept and use human–computer interface systems without discomfort.
Author: 허정
Language: English
Open Access status: green
OpenAlex ID: W2906639839
Cites: 84
Cited by:
Related to: 10