This repository contains the augmented summaries of How2-2000 dataset. We declare that the summaries have been generated using OpenAI's ChatGPT3.5Turbo model.
This repository contains content generated by ChatGPT, a language model developed by OpenAI.
Users are responsible for reviewing and validating the generated content before using it for any purpose. The creators and maintainers of this repository are not liable for any inaccuracies, errors, or consequences arising from the use of ChatGPT-generated content.
Generated summaries used in our experiments are located as below.
direct
corresponds to direct AugSumm
in the paper.
paraphrase
corresponds to paraphrase AugSumm
in the paper using concept words.
./
`-- augsumm
|-- direct
| |-- dev5_test_summaries
| `-- tr_2000h_summaries
`-- paraphrase
|-- dev5_test_summaries
`-- tr_2000h_summaries
All training has been done using ESPnet-SUMM. Please refer to ESPnet and ESPnet-SUMM.
- AugSumm
@inproceedings{jung2024augsumm,
title={Clotho: An audio captioning dataset},
author={Jung, Jee-weon and Sharma, Roshan and Chen, William and Raj, Bhiksha and Watanabe, Shinji},
booktitle={Proc. IEEE ICASSP},
year={2024},
}
- ESPnet
@inproceedings{watanabe2018espnet,
title={ESPnet: End-to-End Speech Processing Toolkit},
author={Watanabe, Shinji and Hori, Takaaki and Karita, Shigeki and Hayashi, Tomoki and Nishitoba, Jiro and Unno, Yuya and Enrique Yalta Soplin, Nelson and Heymann, Jahn and Wiesner, Matthew and Chen, Nanxin and others},
booktitle={Proc. Interspeech},
year={2018}
}
- ESPnet-SUMM
@inproceedings{sharma2023espnetsumm,
title={ESPNet-SUMM: Introducing a novel large dataset, toolkit, and a cross-corpora evaluation of speech summarization systems},
author={Sharma, Roshan and Chen, William and Kano, Takatomo and Sharma, Ruchira and Ogawa, Atsunori and Delcroix, Marc and Watanabe, Shinji and Singh, Rita and Raj, Bhiksha Raj},
booktitle={Proc. IEEE ASRU},
year={2013}
}