| dataset | data type | size | with audio sources? |
|---|---|---|---|
| Amg1608 | frame-level acoustic features and corresponding valence & arousal annotations | 1608 excerpts (30s) | no |
| DEAM | Continuous (per second) and overall valence & arousal ratings | 1802 excerpts | yes |
| DEAPDataset | valence & arousal & dominance & physiological data | 120 music video excerpts | no |
| EMO-Soundscapes | arousal & valence annotations of soundscapes | 1213 soundscape recordings | yes |
| emoMusic | arousal & valence annotations of music | 744 excerpts (45s) | yes |
| Emotify | induced musical emotion from a game using the GEMS scale | 400 excerpts | yes |
| EMusic | arousal & valence | 100 excerpts (experimental music) | yes |
| Greek Music Dataset | genre & valence & arousal with audio features and metadata for popular Greek tracks | 1400 songs | downloadable |
| IADS | normative ratings of emotion (pleasure, arousal, dominance) for a set of acoustic stimuli | 111 sound snippets | yes |
| MOODetector:Bi-Modal | lyrics and valence & arousal annotations | 133 excerpts | yes |
| MOODetector:Multi-Modal | lyrics & MIDI & mood | 903 excerpts (30s) | yes |
| moodswings | arousal & valence | 240 excerpts (30s) | no |
| NMED-H | EEG | 24 trials x 16 excerpts (4.5min) | no |
| NMED-RP | EEG | 20 trials x 10 excerpts (4.5min) | no |
| RAVDESS | 7356 pieces, each file rated 10 times on emotional validity, intensity, and genuineness | 7356 files - 2880 speech files, 2024 song files | yes |
| VGMIDI Dataset | 95 MIDI labelled piano pieces using valence-arousal model of emotion | 95 labelled, 728 non-labelled | yes |
| YouTube Audio Library | royalty-free/creative commons liscened audio with mood tagging and audio metadata | 790 songs | yes |
Created
November 15, 2019 05:33
-
-
Save lwlsn/3621dd65a25a8a342cb99ca673a4a100 to your computer and use it in GitHub Desktop.
Music and emotion datasets
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment