| import os | |
| from typing import Union | |
| import datasets | |
| _DESCRIPTION = """\ | |
| Dusha is a bi-modal corpus suitable for speech emotion recognition (SER) tasks. | |
| The dataset consists of audio recordings with Russian speech and their emotional labels. | |
| The corpus contains approximately 350 hours of data. Four basic emotions that usually appear in a dialog with | |
| a virtual assistant were selected: Happiness (Positive), Sadness, Anger and Neutral emotion. | |
| """ | |
| _HOMEPAGE = "https://github.com/salute-developers/golos/tree/master/dusha#dusha-dataset" | |
| _DATA_URL = "https://huggingface.co/datasets/KELONMYOSA/dusha_emotion_audio/resolve/main/data.zip" | |
| _METADATA_URL = "https://huggingface.co/datasets/KELONMYOSA/dusha_emotion_audio/resolve/main/data.csv" | |
| class Dusha(datasets.GeneratorBasedBuilder): | |
| DEFAULT_WRITER_BATCH_SIZE = 256 | |
| def _info(self): | |
| return datasets.DatasetInfo( | |
| description=_DESCRIPTION, | |
| features=datasets.Features( | |
| { | |
| "path": datasets.Value("string"), | |
| "audio": datasets.Audio(sampling_rate=16_000), | |
| "label": datasets.Value("string"), | |
| } | |
| ), | |
| supervised_keys=None, | |
| homepage=_HOMEPAGE, | |
| ) | |
| def _split_generators(self, dl_manager): | |
| metadata = dl_manager.download(_METADATA_URL) | |
| archive = dl_manager.download(_DATA_URL) | |
| return [datasets.SplitGenerator( | |
| name=datasets.Split.TRAIN, | |
| gen_kwargs={ | |
| "audio_files": dl_manager.iter_archive(archive), | |
| "metadata": metadata}, | |
| )] | |
| def _generate_examples(self, audio_files: Union[str, os.PathLike], metadata: str): | |
| examples = dict() | |
| with open(metadata) as f: | |
| for row in f: | |
| data = row.split(",") | |
| audio_path = data[0] | |
| label = data[1] | |
| res = dict() | |
| res["path"] = os.path.join(audio_files, audio_path) | |
| res["label"] = label | |
| examples[audio_path] = res | |
| key = 0 | |
| for path, f in audio_files: | |
| if path in examples: | |
| audio = {"path": path, "bytes": f.read()} | |
| yield key, {**examples[path], "audio": audio} | |
| key += 1 | |