Dynamic Music emotion recognition based on CNN-BiLSTM
Pengfei Du, Xiaoyong Li, Yali Gao
Abstract
Most of the music information will express the emotional signal, the emotional classification of music can be more convenient for the organization and retrieval of music, is the use of music for psychological intervention and physiological adjustment of the premise. In this paper, the dynamic emotion recognition of music, using the VA model, a song to produce a set of VA value. This paper first elaborates the relevant background and emotion classification model of music emotion recognition, and explain the representations of emotion, designs the overall framework of dynamic music emotion recognition, and adopts the convolution neural network and the BiLSTM fusion way to carry on the music dynamic VA And then compared with other relevant methods of identification.