چكيده به لاتين
In recent decades, the use of electroencephalogram (EEG) signals to communicate with the surrounding environment has led to the advent of brain-computer interfaces (BCIs). The P300 signal is a type of brain potentials that is used in the BCI systems. In this study, the visual stimulation is used to stimulate the P300 signal. Accordingly, three graphical user interfaces (GUIs) are designed for the main menu, the TV control, and the telephone; each of them contains a different number of keys. In these designs, we have attempted to use the innovative GUI to increase the accuracy and decrease the false positive rate of the distinguishment between the target and the non-target items. In addition, users are able to select and control one of the house devices from the main menu GUI. For example, they can change the volume, the TV channel, or select the telephone and dial a number. Moreover, a brain switch was designed for the system to allow the users to change the system state and switch between the control and the idle state whenever they desire. To separate the target and the non-target items and to separate the control and the idle state, we have combined the Spatial-Temporal Discriminant Analysis (STDA) classifying algorithm with the Stacked Auto-Encoder (SAE) analysis. In the online experiments, the subjects have carried out two tasks; whereupon, the classification accuracies were 80.8% and 80.6%; and the Information Transfer Rates (ITRs) were 27.6 and 19.7 bits per minute, respectively.
