Dataset Card for Whisper Tokusatsu

Dataset Description

This dataset contains audio segments from Japanese Tokusatsu series with their corresponding transcriptions.

Languages

The dataset is in Japanese (ja).

Dataset Structure

The dataset contains audio files (.wav) and their transcriptions.

Data Instances

An example of a data instance:

{
    'audio': {
        'path': 'path/to/audio/file.wav',
        'array': [...],  # The actual audio time series
        'sampling_rate': 44100
    },
    'text': '音声のテキスト転写がここに入ります。'  # The transcription of the audio
}

Data Fields

audio: The audio file containing the spoken content.
text: The transcription of the spoken content in the audio file.

Data Splits

The dataset doesn't have predefined splits.

Dataset Creation

Source Data

The audio segments are extracted from Tokusatsu series episodes.

Annotations

The transcriptions are based on the subtitle files of the episodes.

Considerations for Using the Data

Social Impact of Dataset

This dataset can be used to improve speech recognition models for Japanese, particularly in the context of Tokusatsu media.

Discussion of Biases

The dataset may contain biases inherent in the source material, including gender representation and language specific to the Tokusatsu genre.

Other Known Limitations

The quality of transcriptions may vary based on the accuracy of the original subtitle files.

Additional Information

Dataset Curators

This dataset was curated by [Your Name/Organization].

Licensing Information

This dataset is distributed under the MIT License.

Citation Information

If you use this dataset in your research, please cite it as follows:

@dataset{whisper_tokusatsu,
  author    = {[Your Name]},
  title     = {Whisper Tokusatsu Dataset},
  year      = {2024},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/NekoFi/whisper_toku}
}

Contributions

Thanks to [@NekoFi] for adding this dataset.

Visualization

Here's a visualization of the data:

Audio	Text
	おめでとうございます｡
	今日からあなたは仮面ライダーです｡

audio audioduration (s) 1.77 8