The research on cross-platform transplantation of generic model on subject-independent BCI

With the development of wearable technology, it is gradually general to collect EEG signals in portable system with proprietary wireless. However, the introduction of wearable signal collection device always means the descent of EEG quality, sometimes with the invalid of discrimination model in Brain-Computer Interface (BCI). This paper aims at the differences of EEG features from P300-Speller in diverse EEG collection platforms, discussing the transplantation of generic model from high signal noise ratio (SNR) EEG collection platform to the lower one. Results indicate that it is valid for generic model transplantation and model adjustment strategy promotes identification results effectively. It provides technology support for further development of cross-platform BCI.