16 channel Three speaker dynamic switch cEEGrid Auditory Attention Decoding Dataset Nanjing University

An ear-EEG Auditory Attention Decoding Dataset was presented. The 16 channel cEEGrid signal of 98 participants were recorded. Participants attended to three concurrent competing speakers, and were required to attended to one of them. Additionally, participants were required to alter their attention to another speaker, or to ignore all speakers, following instructions in the screen, during the middle of a trial. Each participant took 63 trials, 30 seconds each, resulting in 31.5 minutes of cEEGrid recordings. The ear-EEG signal was captured by OpenBCI CytonDaisy Board with a sampling rate of 125 Hz. Audio stimuli comprised of professional Mandarin Chinese anchor.