Using a dataset of annotated music examples, we train recurrent neural networks to attempt to predict emotion values (valence and arousal) directly from audio waveforms.
Senior
Loading Thumbnail...
?