The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

L2-ARCTIC v5.0

L2-ARCTIC is a non-native English speech corpus intended for research in pronunciation assessment, accent conversion, voice conversion, and mispronunciation detection.

This Hub dataset mirrors the v5.0 release as distributed locally: speaker-level zip archives, the suitcase corpus archive, prompts, the original README, and the license.

Summary

24 non-native English speakers
26,867 utterances
27.1 hours of scripted speech
Manual phone-level annotations for 3,599 utterances
Additional suitcase_corpus spontaneous-speech subset
Native language groups: Arabic, Chinese, Hindi, Korean, Spanish, Vietnamese

Structure

Each speaker archive contains:

wav/: 44.1 kHz WAV audio
transcript/: orthographic transcripts
textgrid/: forced-aligned phoneme transcriptions
annotation/: manual pronunciation annotations where available

License

This dataset is distributed under CC BY-NC 4.0 as provided in the original release. Redistribution is for non-commercial use only.

Citation

@inproceedings{zhao2018l2arctic,
  author={Guanlong {Zhao} and Sinem {Sonsaat} and Alif {Silpachai}
          and Ivana {Lucic} and Evgeny {Chukharev-Hudilainen}
          and John {Levis} and Ricardo {Gutierrez-Osuna}},
  title={L2-ARCTIC: A Non-native English Speech Corpus},
  year=2018,
  booktitle={Proc. Interspeech},
  pages={2783--2787},
  doi={10.21437/Interspeech.2018-1110}
}

Notes

This card is a Hub-specific wrapper around the original release metadata.
Refer to README.original.md for the full upstream release notes.
The original LICENSE file is included unchanged.

Downloads last month: 10