講演情報
[5O3-IS-5b-01]Proposal for creating an EEG dataset of users listening to and recalling piano pitches and chords for machine learning
〇Yuki Kanda1, Motoyuki Sanada2, Yasushi Naruse2, Ikuko Eguchi Yairi1 (1. Sophia University, 2. National Institute of Information and Communications Technology)
regular
キーワード：
EEG、ERP、music、pitch、recall
This paper aims to measure and analyze electroencephalogram (EEG) signals during the listening and recall of single notes and chords using a piano sound source, while also creating an EEG dataset for machine learning. The EEG measurement experiment involved 8 adults with normal hearing. They listened to and recalled 13 types of single notes (C4 to C5) and 12 types of chords (Major, Minor, sus4, and aug chords with root notes C, F, and B♭). Event-related potentials (ERPs) were calculated for both the listening condition and the subsequent silent period during the recall condition. In the listening condition, N100 and P200 were observed at all electrodes, confirming the stability of initial auditory processing. Significant differences in P200 amplitude and components after 200ms were observed depending on the sound type. In contrast, the recall condition showed significantly more differences between pitches and chords than the listening condition.