Research Article

High-Frequency Vibrating Stimuli Using the Low-Cost Coin-Type Motors for SSSEP-Based BCI

Figure 3

An illustration of the SSSEP data processing sequence. The discriminative features were extracted using the FBCSP and were selected using the MIBIF algorithm. Selected features were classified using an SVM classifier. (a) is a shortened process for the CSP data process with a single band-pass and a single spatial filter. (b) is an additional 7 frequency band-pass filter for FBCSP and (c) sequential MIBIF feature selection method.