wake.py

# coding=utf-8
# Copyright 2023 The BizzAI and HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#


import csv
import os

import datasets

logger = datasets.logging.get_logger(__name__)


""" BizzBuddy AI Dataset"""

_CITATION = """\
@article{gerz2021multilingual,
  title={Wake word data for Voice assistant trigger in English from spoken data},
  author={Ahmed, Nicholas},
  year={2023}
}
"""

_DESCRIPTION = """\
Wake is training and evaluation resource for wake word
detection task with spoken data. It covers the wake and not wake
intents collected from a multiple participants who agreed to contribute to the development 
of the system on the wake word and the not wake words is a subset of the common voice and speech commands dataset.
"""

_ALL_CONFIGS = sorted([
    "en-US"
])


_DESCRIPTION = "Wake is a dataset for the wake word detection task with spoken data."


_DATA_URL = 'https://huggingface.co/datasets/Ahmed-ibn-Harun/wake-w/resolve/main/data.tar.gz'


class WakeConfig(datasets.BuilderConfig):
    """BuilderConfig for xtreme-s"""

    def __init__(
        self, name, description, data_url
    ):
        super(WakeConfig, self).__init__(
            name=self.name,
            version=datasets.Version("1.0.0", ""),
            description=self.description,
        )
        self.name = name
        self.description = description
        self.data_url = data_url


def _build_config(name):
    return WakeConfig(
        name=name,
        description=_DESCRIPTION,
        data_url=_DATA_URL,
    )


class Wake(datasets.GeneratorBasedBuilder):

    DEFAULT_WRITER_BATCH_SIZE = 1000
    BUILDER_CONFIGS = [_build_config(name) for name in _ALL_CONFIGS + ["all"]]

    def _info(self):
        task_templates = None
        langs = _ALL_CONFIGS
        features = datasets.Features(
            {
                "path": datasets.Value("string"),
                "audio": datasets.Audio(sampling_rate=8_000),
                "wake": datasets.ClassLabel(
                    names=[
                        0,
                        1,
                    ]
                ),
                "lang_id": datasets.ClassLabel(names=langs),
            }
        )

        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            supervised_keys=("audio", "transcription"),
            citation=_CITATION,
            task_templates=task_templates,
        )

    def _split_generators(self, dl_manager):
        langs = (
            _ALL_CONFIGS
            if self.config.name == "all"
            else [self.config.name]
        )

        archive_path = dl_manager.download_and_extract(self.config.data_url)
        audio_path = dl_manager.extract(
            os.path.join(archive_path,  "audio.tar.gz")
        )
        text_path = dl_manager.extract(
            os.path.join(archive_path, "text.tar.gz")
        )

        text_path = {l: os.path.join(text_path, f"{l}.csv") for l in langs}

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "audio_path": audio_path,
                    "text_paths": text_path,
                },
            )
        ]


    def _generate_examples(self, audio_path, text_paths):
        key = 0
        for lang in text_paths.keys():
            text_path = text_paths[lang]
            with open(text_path, encoding="utf-8") as csv_file:
                csv_reader = csv.reader(csv_file, delimiter=",", skipinitialspace=True)
                next(csv_reader)
                for row in csv_reader:
                    file_path, intent_class = row

                    file_path = os.path.join(audio_path, *file_path.split("/"))
                    yield key, {
                        "path": file_path,
                        "audio": file_path,
                        "wake": intent_class,
                        "lang_id": _ALL_CONFIGS.index(lang),
                    }
                    key += 1