|
Title:
|
TWO-PITCH ESTIMATION FROM EEG SIGNALS FOR MUSIC-QUERY BY MELODY-RECALL IN THE BRAIN |
|
Author(s):
|
Kyoji Kawagoe, Mao Otsuka, Ryosuke Yamanishi, Yasuhiro Tsubo, Akira Hirabayashi, Haruo Sakuma |
|
ISBN:
|
978-989-8533-56-2 |
|
Editors:
|
Hans Weghorn |
|
Year:
|
2016 |
|
Edition:
|
Single |
|
Keywords:
|
Music information retrieval, Brain wave, Pitch estimation, SVM, K-NN |
|
Type:
|
Full Paper |
|
First Page:
|
179 |
|
Last Page:
|
185 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Many studies in the field of Music Information Retrieval (MIR) have produced several viable applications for music selection. However, because the primary methods for MIR input are typically keywords or melody humming, there is a significant gap between songs that the user is seeking and the MIR results. This paper describes a novel method of pitch estimation based on melody recall from brain data obtained with EEG sensors. The evaluation results show that our proposed two-pitch method can estimate brain data pitches with an average accuracy of 69%. Our ongoing study is focusing on estimation of three pitches and improving the method for achieving more accurate pitch recognition. |
|
|
|
|
|
|
