DCASE2020 task4: Sound event detection in domestic environments using source separation

• Information about the DCASE 2020 challenge please visit the challenge website.
• You can find discussion about the dcase challenge here: dcase-discussions.

Dependencies

Python >= 3.6, pytorch >= 1.0, cudatoolkit>=9.0, pandas >= 0.24.1, scipy >= 1.2.1, pysoundfile >= 0.10.2, scaper >= 1.3.5, librosa >= 0.6.3, youtube-dl >= 2019.4.30, tqdm >= 4.31.1, ffmpeg >= 4.1, dcase_util >= 0.2.5, sed-eval >= 0.2.1, psds-eval >= 0.1.0, desed >= 1.1.7

Environment

launch conda_create_environment.sh (recommended line by line)

Reproducing the results

See GL-MT folder.

Dataset

Scripts to generate the dataset

In the scripts/ folder, you can find the different steps to:

• Download recorded data and synthetic material.
• Generate synthetic soundscapes

It is likely that you'll have download issues with the real recordings. At the end of the download, please send a mail with the TSV files created in the missing_files directory. (to Nicolas Turpault and Romain Serizel).

**Please store data in the given structure as stated in scripts/

Description

• The sound event detection dataset is using desed dataset.

dataset

The dataset for sound event detection of DCASE2020 task 4 is composed of:

• Train:
  • *weak (DESED, recorded, 1 578 files)
  • *unlabel_in_domain (DESED, recorded, 14 412 files)
  • synthetic20/soundscapes (DESED, 2584 files)
• *Validation (DESED, recorded, 1 168 files):
  • test2018 (288 files)
  • eval2018 (880 files)

Baselines dataset

SED baseline

• Train:
  • weak
  • unlabel_in_domain
  • synthetic20/soundscapes
• Validation:
  • validation

Annotation format

Weak annotations

The weak annotations have been verified manually for a small subset of the training set. The weak annotations are provided in a tab separated csv file (.tsv) under the following format:

[filename (string)][tab][event_labels (strings)]

For example:

Y-BJNMHMZDcU_50.000_60.000.wav	Alarm_bell_ringing,Dog

Strong annotations

Synthetic subset and validation set have strong annotations.

The minimum length for an event is 250ms. The minimum duration of the pause between two events from the same class is 150ms. When the silence between two consecutive events from the same class was less than 150ms the events have been merged to a single event. The strong annotations are provided in a tab separated csv file (.tsv) under the following format:

[filename (string)][tab][event onset time in seconds (float)][tab][event offset time in seconds (float)][tab][event_label (strings)]

For example:

YOTsn73eqbfc_10.000_20.000.wav	0.163	0.665	Alarm_bell_ringing

Authors

Author	Affiliation
Hao Yen	National Taiwan University
Pin-Jui Ku	National Taiwan University

Contact

If you have any problem feel free to contact Hao Yen (b05901090@ntu.edu.tw) or Pin-Jui Ku (b05901107@ntu.edu.tw)