Stacked Convolutional and Recurrent Neural Networks for Music Emotion Recognition
This work addresses emotion recognition from music for applications like recommendation systems, but it is incremental as it builds on existing CNN and RNN techniques.
The paper tackled music emotion recognition in the valence-arousal space by proposing a CNN-RNN hybrid method with fewer parameters than state-of-the-art approaches, achieving RMSE scores of 0.202 for arousal and 0.268 for valence, which are the best reported results on the MediaEval2015 dataset.
This paper studies the emotion recognition from musical tracks in the 2-dimensional valence-arousal (V-A) emotional space. We propose a method based on convolutional (CNN) and recurrent neural networks (RNN), having significantly fewer parameters compared with the state-of-the-art method for the same task. We utilize one CNN layer followed by two branches of RNNs trained separately for arousal and valence. The method was evaluated using the 'MediaEval2015 emotion in music' dataset. We achieved an RMSE of 0.202 for arousal and 0.268 for valence, which is the best result reported on this dataset.