Brain Computer Interface Motor Imagery-EEG Dataset

Brain-computer interface (BCI) is an effective approach for users to control external software applications and devices only by decoding their brain activities and without engaging any muscles. The availability of a BCI dataset which are large-scale and high quality can stimulate the researchers from neighbouring research areas develop advanced deep learning algorithms to enrich the field of BCI. Therefore, it is necessary to build an EEG dataset with availability for the development and research of BCI system. This is a large and high-performance intuitive dataset from World Robot Conference Contest-BCI Robot Contest MI of upper limbs or upper and lower limbs with three recording sessions. Sixty-three healthy participants (all right-handed, aged 17-30, 18 females) were naive BCI users participated in this experiment, of which fifty-two subjects participated in the two-class MI experiment and eleven subjects participated in the three-class MI experiment. This dataset can be utilized for a wide range of BCI-related research questions: analysing the differences among recording sessions of the same subject, improving the decoding performance of algorithms.