trexwithoutt / MusicStyleTransfer

This project is aiming to transfer the style of music data

Music Style Transfer via VAE

Author: Yubi Chen, Rex Zhou

Dependencies

- Python 3.6
- Numpy 1.14
- Keras 2.0.8
- ReccurentShop
- sklearn
- matplotlib

Files

.
├── data_folder_settings.py
├── credentials.json
├── instrument_classifier.ipynb
├── velocity_classifier.ipynb
├── pitch_classifier.ipynb
├── LICENSE.md
├── settings.py
├── token.pickle
├── vae_definition.py
├── README.md
├── vae_evaluation.ipynb
└── vae_training.ipynb

Usage

vae_training.ipynb will train the vae model, and vae_evaluation.ipynb will give different evaluate method for the transfer work.

Reference

• Gino Brunner MIDI-VAE: Modeling Dynamics And Instrumentation of Music With Application to Style Transfer. ISMIR 2018, Paris, France
• Ramon Lopez de Mantaras and Josep Lluis Arcos Ai and music from composition to expressive performance. AI Mag., 23(3):4357, September 2002. ISSN 0738-4602.
• Midi association, the official midi specifications. https://www.midi.org/specifications. Accessed: 01-06-2018.
• Leon A Gatys, Alexander S Ecker, and Matthias Bethge. Image style transfer using convolutional neural networks. In Computer Vision and Pattern Recognition (CVPR), 2016 IEEE Conference on, pages 2414–2423. IEEE, 2016
• A¨aron van den Oord, Oriol Vinyals, and Koray Kavukcuoglu. Neural discrete representation learning. In Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, 4-9 December 2017, Long Beach, CA, USA, pages 6309–6318, 2017