Datasets:
license: apache-2.0
task_categories:
- automatic-speech-recognition
language:
- uz
pretty_name: Uzbek Language Speech-to-Text Dataset
size_categories:
- 100K<n<1M
Uzbek Speech-to-Text Dataset (STT_uz) Dataset Description The Uzbek Speech-to-Text (STT_uz) dataset is designed to facilitate the development and evaluation of speech-to-text models for the Uzbek language. This dataset includes a variety of audio recordings paired with their transcriptions, collected from different sources to ensure diverse coverage of spoken Uzbek.
Dataset Summary Languages: Uzbek Size: Varied across different subsets. Total Audio Duration: Varied File Formats: .wav and mp3 for audio files, .tsv for transcriptions. Dataset Structure The dataset is organized into the following directories and files:
audio/
uz_other_dataset/: Contains .tar archives like uz_other_0.taruz_other_1.tar uz_train_dataset/: Contains .tar archives like uz_train_0.tar. uz_validated_dataset/: Contains .tar archives like uz_validated_0.tar, uz_validated_1.tar, and uz_validated_2.tar. wavs/: Contains individual .wav files. transcription/: Contains .tsv files including:
other.tsv train.tsv validated.tsv wavs.tsv The .tsv files have two columns: file_name and transcription. Each entry provides the path to the audio file and its corresponding transcription.
Data Instances A typical data point includes:
Audio File: Path to the .wav and mp3 files. Transcription: Text transcribed from the audio. Example data instance:
json Copy code { "file_name": "audio/uz_train_dataset/common_voice_uz_28907218.mp3", "transcription": "Bugun ertalab Gyotenikiga taklifnoma oldim." } Data Splits The dataset is divided into the following subsets:
Train: Used for training models. Validated: Used for validation and model tuning. Other: Additional data that has not yet been reviewed. The data was collected from various sources to ensure a diverse representation of spoken Uzbek.
Annotations Annotations were performed manually, ensuring high accuracy of transcriptions.
Considerations for Using the Data Personal and Sensitive Information: The dataset contains voice recordings from contributors who have voluntarily shared their data. Do not attempt to identify any individual contributors. Social Impact: This dataset is intended to improve speech recognition technology for the Uzbek language, with potential applications in various domains. Licensing Information This dataset is licensed under the CC0 1.0 Universal (CC0 1.0) Public Domain Dedication.
Citation Information If you use this dataset in your research or projects, please cite it as follows:
bibtex Copy code @misc{uzbek_stt_dataset, author = {Beehzod}, title = {Uzbek Speech-to-Text Dataset (STT_uz)}, year = {2024}, url = {https://huggingface.co/datasets/Beehzod/STT_uz} }