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.
SHA-1: EF49EAE3C416FEA684F5EFE290A230475454DF7E
Detailed descriptions about the dataset can be found in the dataset description page.
 |
|
Graphical Interface For Annotation
|
Distribution of Average Annotations
|
