Skip to content

source code of "End-to-end Music Remastering System Using Self-supervised and Adversarial Training"

License

Notifications You must be signed in to change notification settings

jhtonyKoo/e2e_music_remastering_system

Repository files navigation

End-to-end Music Remastering System

This repository includes source code and pre-trained models of the work End-to-end Music Remastering System Using Self-supervised and Adversarial Training by Junghyun Koo, Seungryeol Paik, and Kyogu Lee.

We provide inference code of the proposed system, which targets to alter the mastering style of a song to desired reference track.

arXiv Demo Page

Pre-trained Models

Model Number of Epochs Trained Details
Music Effects Encoder 1000 Trained with MTG-Jamendo Dataset
Mastering Cloner 1000 Trained with the above pre-trained Music Effects Encoder and Projection Discriminator

Inference

To run the inference code,

  1. Download pre-trained models above and place them under the folder named 'model_checkpoints' (default)
  2. Prepare input and reference tracks under the folder named 'inference_samples' (default).
    Target files should be organized as follow:
    "path_to_data_directory"/"song_name_#1"/input.wav
    "path_to_data_directory"/"song_name_#1"/reference.wav
    ...
    "path_to_data_directory"/"song_name_#n"/input.wav
    "path_to_data_directory"/"song_name_#n"/reference.wav
  1. Run 'inference.py'
python inference.py \
    --ckpt_dir "path_to_checkpoint_directory" \
    --data_dir_test "path_to_directory_containing_inference_samples"
  1. Outputs will be stored under the folder 'inference_samples' (default)

Note: The system accepts WAV files of stereo-channeled, 44.1kHZ, and 16-bit rate. Target files shold be named "input.wav" and "reference.wav".

Configurations of each sub-networks

config_table

A detailed configuration of each sub-networks can also be found at

Self_Supervised_Music_Remastering_System/configs.yaml

About

source code of "End-to-end Music Remastering System Using Self-supervised and Adversarial Training"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages