EEG Thinking1 Datasets: Think-Count-Recall (TCR) and Read-Write-Type (RWT)

Xiaodong Qu, Peiyan Liu

Keywords: dataset, EEG, machine learning, pipeline, user training, feedback

Abstract: EEG-based Brain-Computer Interfaces (BCI) have been widely used in clinical and non-clinical research. In this paper, we present a framework to collect a large amount of EEG data with an easy-to-use experiment setup, using non-invasive, wireless, and affordable hardware. Interpretable feedback generated by benchmark machine learning algorithms has been provided to the researchers and end-users. Two existing datasets are used as case studies for the framework: Read-Write-Type (RWT) and Think-Count-Recall (TCR). The goal is to inspire new machine learning approaches for decoding behavior from large-scale EEG data. The framework of experimental design, data collection, data analysis, feedback generation, and community building could pave the way towards a future when everyone can easily use BCI systems every day, similar to smartphones nowadays.