Abstract | ||
---|---|---|
Automatic melody generation has been a long-time aspiration for both AI researchers and musicians. However, learning to generate euphonious melodies has turned out to be highly challenging. This paper introduces 1) a new variant of variational autoencoder (VAE), where the model structure is designed in a modularized manner in order to model polyphonic and dynamic music with domain knowledge, and 2) a hierarchical encoding/decoding strategy, which explicitly models the dependency between melodic features. The proposed framework is capable of generating distinct melodies that sounds natural, and the experiments for evaluating generated music clips show that the proposed model outperforms the baselines in human evaluation.
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/icassp.2019.8683106 | ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) |
Keywords | Field | DocType |
Music Generation,VAE,Modularization | Melody,Autoencoder,Domain knowledge,Dynamics (music),Computer science,Speech recognition,Artificial intelligence,Polyphony,Decoding methods,Machine learning,Feature dependency,Encoding (memory) | Journal |
Volume | ISSN | Citations |
abs/1811.00162 | 1520-6149 | 0 |
PageRank | References | Authors |
0.34 | 0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yu-An Wang | 1 | 0 | 0.68 |
Yu-Kai Huang | 2 | 0 | 0.34 |
Tzu-Chuan Lin | 3 | 0 | 0.34 |
Shang-Yu Su | 4 | 9 | 4.88 |
Yun-Nung Chen | 5 | 324 | 35.41 |