@inproceedings{liu-etal-2020-adapting,
    title = "Adapting End-to-End Speech Recognition for Readable Subtitles",
    author = "Liu, Danni  and
      Niehues, Jan  and
      Spanakis, Gerasimos",
    editor = {Federico, Marcello  and
      Waibel, Alex  and
      Knight, Kevin  and
      Nakamura, Satoshi  and
      Ney, Hermann  and
      Niehues, Jan  and
      St{\"u}ker, Sebastian  and
      Wu, Dekai  and
      Mariani, Joseph  and
      Yvon, Francois},
    booktitle = "Proceedings of the 17th International Conference on Spoken Language Translation",
    month = jul,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/2020.iwslt-1.30/",
    doi = "10.18653/v1/2020.iwslt-1.30",
    pages = "247--256",
    abstract = "Automatic speech recognition (ASR) systems are primarily evaluated on transcription accuracy. However, in some use cases such as subtitling, verbatim transcription would reduce output readability given limited screen size and reading time. Therefore, this work focuses on ASR with output compression, a task challenging for supervised approaches due to the scarcity of training data. We first investigate a cascaded system, where an unsupervised compression model is used to post-edit the transcribed speech. We then compare several methods of end-to-end speech recognition under output length constraints. The experiments show that with limited data far less than needed for training a model from scratch, we can adapt a Transformer-based ASR model to incorporate both transcription and compression capabilities. Furthermore, the best performance in terms of WER and ROUGE scores is achieved by explicitly modeling the length constraints within the end-to-end ASR system."
}