DnR-v2 / README.md
Den4ikAI's picture
Add dataset card (mirrored from Zenodo)
d9bd380 verified
|
Raw History Blame Contribute Delete
6.14 kB
metadata
license: cc-by-4.0
pretty_name: Divide and Remaster (DnR) v2
task_categories:
  - audio-to-audio
  - audio-classification
  - automatic-speech-recognition
tags:
  - audio
  - audio source separation
  - soundtrack separation
  - cinematic audio source separation
  - sound event detection
  - music genre recognition
  - speech recognition
size_categories:
  - 1K<n<10K

Divide and Remaster (DnR) v2

This is an unofficial mirror. The dataset was created by Darius Petermann, Gordon Wichern, Zhong-Qiu Wang and Jonathan Le Roux (Indiana University / Mitsubishi Electric Research Laboratories) and originally published on Zenodo: https://zenodo.org/records/6949108 (DOI: 10.5281/zenodo.6949108).

The files here are byte-identical copies of the Zenodo files (same MD5 checksums, listed below). Nothing has been repackaged or modified. The mirror is hosted only to make downloads faster. I claim no rights to the data. All credit goes to the original authors. The dataset is distributed under the original CC BY 4.0 license. Please cite the original paper (see Citation).

The description below is copied from the original Zenodo record.

Introduction

Divide and Remaster (DnR) is a source separation dataset for training and testing algorithms that separate a monaural audio signal into speech, music, and sound effects/background stems. The dataset is composed of artificial mixtures using audio from the librispeech, free music archive (FMA), and Freesound Dataset 50k (FSD50k). We introduce it as part of the Cocktail Fork Problem paper.

At a Glance

  • The size of the unzipped dataset is ~200GB
  • Each mixture is 60 seconds long and sources are not fully overlapped
  • Audio is encoded as 32-bit .wav files at a sampling rate of 44.1 kHz
  • The data is split into training tr (3406 mixtues), validation cv (487 mixtures) and testing tt (973 mixtures) subsets
  • The directory for each mixture contains four .wav files, mix.wav, music.wav, speech.wav, sfx.wav, and annots.csv which contains the metadata for the original audio used to compose the mixture (transcriptions for speech, sound classes for sfx, and genre labels for music)

Other Resources

Demo examples and additional information are available at: https://cocktail-fork.github.io/

For more details about the data generation process, the code used to generate our dataset can be found at the following: https://github.com/darius522/dnr-utils

Fix Brought to V2

V1 contained some errors as for the speech annotations, which in some cases did not match with their associated audio utterances. This was caused by some of the speech utterances being truncated to fit the length of their associated DnR mixtures, while their transcriptions were reported in their entirety. To address the issue, we discarded the possibility of any utterances being truncated during the DnR creation process. Since the number of test-set mixtures is determined such that we exhaust all utterances from the LibriSpeech TEST-CLEAN set twice, the test-set grew larger in V2 (from 652 to 973 mixtures). To maintain the same split proportions, the training and validation sets have been increased accordingly as well (3406 and 487, respectively).

All the results in the camera-ready paper have also been updated to reflect these changes and using the DnR V2.

In this version, we split the datasets into smaller chunks to ease-up the download process.

Contact and Support

Have an issue, concern, or question about DnR? If so, please open an issue here.

(Questions about this mirror itself can go to the Community tab of this repository.)

Download

DnR V2 is split into smaller ~10GB chunks to ease-up the download process. First download all the chunks into a single directory:

hf download Den4ikAI/DnR-v2 --repo-type dataset --local-dir dnr_v2
cd dnr_v2

Files:

File Size (bytes) MD5
dnr_v2.tar.gz.00 10737418240 c4d200e054e2b80b82bdaf70bd9927a2
dnr_v2.tar.gz.01 10737418240 f5ad60624a69c1760de6289a0e4f391a
dnr_v2.tar.gz.02 10737418240 41cd14f12b90a6347afac1e3fad8506a
dnr_v2.tar.gz.03 10737418240 cf5ec720d68f90ef461fe1713118913b
dnr_v2.tar.gz.04 10737418240 8f56a37b39ce65df6aca66fc07159d77
dnr_v2.tar.gz.05 10737418240 b6d4bc7a2592bf09e5985beb2492cbd9
dnr_v2.tar.gz.06 10737418240 23e9ce5861fa33b1667ba303cf759e0e
dnr_v2.tar.gz.07 10737418240 8e8c115fcd083bc4215e51e36709e6a1
dnr_v2.tar.gz.08 10737418240 8c3d65ebf197abb5054edfc8e6f77eb8
dnr_v2.tar.gz.09 10737418240 eabe1367b91c12c3ccb9d160ddcf2aff
dnr_v2.tar.gz.10 9046014144 e06ae1858761341365579de3a1228dc1

From the same directory run the following to compile all the chunks into a single .tar file:

cat dnr_v2.tar.gz.* >dnr_v2.tar.gz

Finally untar the resulting file:

tar -xf dnr_v2.tar.gz

License

Creative Commons Attribution 4.0 International (CC BY 4.0), as set by the original authors on Zenodo. The mixtures are built from LibriSpeech, the Free Music Archive (FMA) and FSD50K. See the original sources for the licenses of the underlying recordings.

Citation

If you use DnR please cite our paper in which we introduce the dataset as part of the Cocktail Fork Problem:

@INPROCEEDINGS{petermann2021cfp,
  author={Petermann, Darius and Wichern, Gordon and Wang, Zhong-Qiu and Roux, Jonathan Le},
  booktitle={ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  title={The Cocktail Fork Problem: Three-Stem Audio Separation for Real-World Soundtracks},
  year={2022},
  volume={},
  number={},
  pages={526-530},
  doi={10.1109/ICASSP43922.2022.9746005}
}

Dataset record: Petermann, D., Wichern, G., Wang, Z.-Q., & Le Roux, J. (2021). Divide and Remaster (DnR) (Version 2.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.6949108