artur-muratov/multilingual-speech-commands-15lang-zip

Multilingual Speech Commands Dataset (15 Languages, Augmented)

This dataset contains augmented speech command samples in 15 languages, derived from multiple public datasets. Only commands that overlap with the Google Speech Commands (GSC) vocabulary are included, making the dataset suitable for multilingual keyword spotting tasks aligned with GSC-style classification.

Audio samples have been augmented using standard audio techniques to improve model robustness (e.g., time-shifting, noise injection, pitch variation).

The dataset is organized as a set of zip files, one per command label (e.g., yes.zip, no.zip, go.zip, stop.zip, etc.). Metadata files are included to facilitate training and evaluation.

Languages

• English
• Kazakh
• Russian
• Tatar
• Arabic
• Turkish
• French
• German
• Spanish
• Italian
• Catalan
• Persian
• Polish
• Dutch
• Kinyarwanda

Structure

The dataset includes:

• One zip file per speech command (e.g., yes.zip, no.zip, go.zip, stop.zip, etc.), each containing the audio samples for that command
• Metadata files:
  • training_list.txt
  • validation_list.txt
  • testing_list.txt
  • label_map.json
  • lang_map.json

License

All audio samples are redistributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license, as permitted by the original sources. Please cite the original datasets below if you use this dataset in your research.

Citations

@article{speechcommandsv2,
author = {{Warden}, P.},
title = "{Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition}",
journal = {ArXiv e-prints},
archivePrefix = "arXiv",
eprint = {1804.03209},
year = 2018,
month = apr,
url = {[https://arxiv.org/abs/1804.03209}](https://arxiv.org/abs/1804.03209})
}

@INPROCEEDINGS{TatarSCR,  
author={Kuzdeuov, Askat and Gilmullin, Rinat and Khakimov, Bulat and Varol, Huseyin Atakan},  
booktitle={Proc. of the Annual Conference of the IEEE Industrial Electronics Society (IECON)},  
title={An Open-Source Tatar Speech Commands Dataset for IoT and Robotics Applications},  
year={2024},  
pages={1-5},  
doi={10.1109/IECON55916.2024.10905876}  
}

@INPROCEEDINGS{KazakhSCR,  
author={Kuzdeuov, Askat and Nurgaliyev, Shakhizat and Turmakhan, Diana and Laiyk, Nurkhan and Varol, Huseyin Atakan},  
booktitle={Proc. of the International Conference on Robotics, Automation and Artificial Intelligence (RAAI)},  
title={Speech Command Recognition: Text-to-Speech and Speech Corpus Scraping Are All You Need},  
year={2023},  
pages={286-291},  
doi={10.1109/RAAI59955.2023.10601292}  
}

@INPROCEEDINGS{MultilingualSCR,  
author={Kuzdeuov, Askat and Varol, Huseyin Atakan},  
booktitle={2025 11th International Conference on Control, Automation and Robotics (ICCAR)},  
title={Multilingual Speech Command Recognition for Voice Controlled Robots and Smart Systems},  
year={2025}  
}

@inproceedings{mazumder2021multilingual,
title = {Multilingual Spoken Words Corpus},
author = {Mazumder, Mark and Chitlangia, Sharad and Banbury, Colby and Kang, Yiping and Ciro, Juan Manuel and Achorn, Keith and Galvez, Daniel and Sabini, Mark and Mattson, Peter and Kanter, David and Diamos, Greg and Warden, Pete and Meyer, Josh and Reddi, Vijay Janapa},
booktitle = {Proc. of the NeurIPS Datasets and Benchmarks Track},
year = {2021},
url = {[https://mlcommons.org/en/multilingual-spoken-words}](https://mlcommons.org/en/multilingual-spoken-words})
}

@article{GHANDOURA2021104267,
author = {Ghandoura, Abdulkader and Hjabo, Farouk and Al Dakkak, Oumayma},
title = {Building and Benchmarking an Arabic Speech Commands Dataset for Small-Footprint Keyword Spotting},
journal = {Engineering Applications of Artificial Intelligence},
year = {2021},
publisher = {Elsevier}
}

@misc{turkish_speech_commands,
author = {Murat Kurtkaya},
title = {Turkish Speech Command Dataset},
year = {2021},
howpublished = {\url{[https://www.kaggle.com/datasets/muratkurtkaya/turkish-speech-command-dataset}}](https://www.kaggle.com/datasets/muratkurtkaya/turkish-speech-command-dataset}})
}

Acknowledgements

This dataset reuses and builds upon data made publicly available by the authors of the original datasets listed above. We thank them for enabling open research.