metadata

license: apache-2.0
language:
  - zh
size:
  - 22020 pairs_of_segments

BSTC-TextRewrite Dataset

Dataset Description

This dataset includes text data that has been annotated for rewriting using the manual annotation method described in our paper to enhance the performance of speech translation systems. It is based on the original BSTC dataset.

Dataset Structure

The dataset consists of pairs of segments structured as follows:

left: The ASR transcript.
right: The human annotation text.

Dataset Statistics

Our dataset contains 22,000 annotation pairs.

License

This dataset is licensed under the Apache 2.0 License. Please ensure you follow the original BSTC dataset license and any applicable terms when using this rewritten dataset.

Citation

If you use this dataset in your work, please cite the following paper:

@article{wu2025improvespeechtranslation,
  title={Improve Speech Translation Through Text Rewrite},
  author={Wu, Jing and Wang, Shushu and Fan, Kai and Luo, Wei and Liao, Minpeng and Huang, Zhongqiang},
  year={2025}
}