rmarcosg/bark-detection

Bark detection dataset

Dataset Description

This dataset comprises both positive and negative samples of audio of 1 second in WAV format, recorded at 44.1kHz.

Negative samples include music, voice, claps, whistles and vacuum cleaner noise, among other sound you may record inside a house.

Caveats:

• This is an imbalanced dataset: ~10k negatives vs ~500 positives.
• Positive samples may include human generated barks.
• Some (few) positive samples are false positives.

Example usage

from datasets import load_dataset
dataset = load_dataset("rmarcosg/bark-detection", split="train")

Supported Tasks

• audio-classification: This dataset was created for the task of binary classification.

Dataset Structure

{
  'file_id': '20221203_185908.wav',
  'audio': {
    'path': 'train/no/20221203_185908.wav',
    'array': array([ 6.10351562e-05, -3.05175781e-05, -3.05175781e-05, ...,
      6.10351562e-05, -1.22070312e-04,  1.83105469e-04]),
    'sampling_rate': 44100
  },
  'label': 'no'
}

Additional Information

Licensing Information

The dataset is available under the Apache 2.0 license.

Citation Information

Please cite the dataset when using it as:

@InProceedings{huggingface:dataset,
title = {A bark detection dataset with positive and negative samples of 1 second},
author={Rodrigo Marcos García},
year={2024}
}

Contributions

Thanks to Greta (my dog) for its help.