Calliope -- A Polyphonic Music Transformer
This work addresses the problem of efficient polyphonic music modeling for AI and music generation applications, representing an incremental advance with specific gains.
The authors tackled the challenge of modeling polyphonic music with deep learning by introducing Calliope, a novel autoencoder based on Transformers, which improved state-of-the-art results in musical sequence reconstruction and generation, particularly on long sequences.
The polyphonic nature of music makes the application of deep learning to music modelling a challenging task. On the other hand, the Transformer architecture seems to be a good fit for this kind of data. In this work, we present Calliope, a novel autoencoder model based on Transformers for the efficient modelling of multi-track sequences of polyphonic music. The experiments show that our model is able to improve the state of the art on musical sequence reconstruction and generation, with remarkably good results especially on long sequences.