Pop Music Highlighter

TensorFlow implementation of Pop Music Highlighter: Marking the Emotion Keypoints

Please cite this paper if this code/work is helpful:

@article{huang2018highlighter,
  title={Pop music highlighter: Marking the emotion keypoints},
  author={Huang, Yu-Siang and Chou, Szu-Yu and Yang, Yi-Hsuan},
  journal={Transactions of the International Society for Music Information Retrieval},
  year={2018},
  volume={1},
  number={1},
  pages={68--78}
}

Environment

Note: you need to rewrite the main.py for your own purpose and the input audio format to be (mp3 format).

$ git clone https://github.com/remyhuang/pop-music-highlighter.git  
$ cd pop-music-highlighter
$ python main.py

Outputs

Three default output files

Possible Error

License

The source code is licensed under GNU General Public License v3.0. However, the pre-trained model (those files under the folder 'model') is licensed under CC BY-NC 4.0. Academia Sinica (Taipei, Taiwan) reserves all the copyrights for the pre-trained model.

Contact

Please feel free to contact Yu-Siang Huang if you have any questions.