The AMG1608
Dataset for Music Emotion Recognition
Introduction
AMG1608 is a dataset for music emotion analysis. It contains
frame-level acoustic features extracted from 1608 30-second music clips and
corresponding valence-arousal (VA) annotations provided by 665 subjects.
The dataset can be
downloaded using this
link.
SHA-1: EF49EAE3C416FEA684F5EFE290A230475454DF7E
Detailed descriptions about the dataset can be found
in the dataset
description page.
|
|
|
|
Graphical
Interface For Annotation |
Distribution
of Average Annotations |
