SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018)
[ACTIVE]
Speech-to-Text for Broadcasters, From Research to Implementation
Metadata
- Publisher
- SMPTE — White Plains, NY, USA
- Doc Type
- Journal Article
- Content Type
- Original Research
- Abbreviated Title
- SMPTE Mot. Imag. J
- Volume
- 127, No. 2, pp. 27–33
- Abstract
- Speech to text has recently moved from the laboratory to the newsroom as a tool for broadcasters and journalists. Breakthroughs in automatic analysis and improvements in affordability mean that running it at a scale of over hundreds of thousands of hours of content is now feasible. Increases in accuracy mean that users will have a realistic chance of finding what they want in minutes rather than hours, especially in genres such as news or factual content. In this paper, we detail the work the British Broadcasting Corp. (BBC) Research and Development (R & D) have done in this area, and in particular how we built a speech-to-text system using open-source tools, broadcast audio, and subtitles. We detail its accuracy across a range of genres and how it performs on real-life broadcasting problems such as cross-talk, music beds, and laughter. We also explain how an in-house innovation team called BBC News Labs took our work and turned it into digital systems that journalists and program makers can use.
- Publication Date
- 2018-03-01
- DOI
10.5594/JMI.2018.2790658- ISSN
- Print:
1545-0279| Electronic:2160-2492 - Link
- https://doi.org/10.5594/JMI.2018.2790658
- Author(s)
- Matt HaynesAlex NortonAndrew McParlandRob Cooper
- Keyword(s)
- Speech-to-Text, Broadcast, Deep Neural Networks, Machine Learning, Open Source, Automatic Speech Recognition, Subtitles
- Copyright
- © 2018 SMPTE
Bibliographic Reference(s)
- “World Service Archive Prototype.” [Online]. Available: http://www.bbc.co.uk/rd/projects/worldservice-archive-proto EXTERNAL
- P. Bell et al., “The MGB Challenge: Evaluating Multi-Genre Broadcast Media Recognition,” 2015. [Online]. Available: http://homepages.inf.ed.ac.uk/srenals/mgb-asru2015.pdf EXTERNAL
- “Tensorflow.” [Online]. Available: https://www.tensorflow.org/ EXTERNAL
- “Lium Diarisation Software.” [Online]. Available: http://www-lium.univ-lemans.fr/diarization/doku.php EXTERNAL
- D. Povey et al., “Generating Exact Lattices in the WST Framework,” 2012. [Online]. Available: http://www.danielpovey.com/files/2012_icassp_lattices.pdf EXTERNAL
- M. Evans, “Speech Recognition in Assisted and Live Subtitling for Television,”2003. [Online]. Available: http://www.bbc.co.uk/rd/publications/whitepaper065 EXTERNAL
- D. Povey et al., “The Kaldi Speech Recognition Toolkit,” 2011. [Online]. Available: http://www.danielpovey.com/files/2011_asru_kaldi.pdf EXTERNAL
- Y. Raimond and F. Ferne, “The BBC World Service Archive Prototype,” 2013. [Online]. Available: http://challenge.semantic-web.org/2013/submissions/swc2013_submission_5.pdf EXTERNAL
- “Natural Speech Technology.” [Online]. Available: http://w w w.research.ed.ac.uk /por tal /en /projects/natural-speechtechnology(3cd7d9a4-ded1-4613-a30b-3fa8a7b21d70).html EXTERNAL
- A. Graves and N. Jaitly, “Toward End-to-End Speech Recognition with Recurrent Neural Networks,” 2014. [Online]. Available: http://proceedings.mlr.press/v32/graves14.pdf EXTERNAL
- X. Zhang, “Improving Deep Neural Network Acoustic Models Using Generalized Maxout Networks,” 2014. [Online]. Available: http://www.danielpovey.com/files/2014_icassp_dnn.pdf EXTERNAL
- M. Mohri, F. Pereira, and M. Riley, “Weighted Finite-State Transducers in Speech Recognition,” 2001. [Online]. Available: http://www.openfst.org/twiki/pub/FST/FstBackground/csl01.pdf EXTERNAL
Source Data (JSON)
Full registry record with provenance metadata. Open directly: /api/doc/10.5594-JMI.2018.2790658.json
Reference Tree
Explore all references and references to this document, as a navigable tree.
Open Reference TreeReference this Doc
Plain text (ISO 690 compliant)
Preview:
Matt Haynes, Alex Norton, Andrew McParland, and Rob Cooper; Speech-to-Text for Broadcasters, From Research to Implementation, SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018); SMPTE, 2018. Available at https://doi.org/10.5594/JMI.2018.2790658
Snippet:
Matt Haynes, Alex Norton, Andrew McParland, and Rob Cooper; Speech-to-Text for Broadcasters, From Research to Implementation, SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018); SMPTE, 2018. Available at https://doi.org/10.5594/JMI.2018.2790658
HTML (ISO 690 compliant)
Preview:
Matt Haynes, Alex Norton, Andrew McParland, and Rob Cooper; Speech-to-Text for Broadcasters, From Research to Implementation, SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018); SMPTE, 2018. Available at https://doi.org/10.5594/JMI.2018.2790658
Snippet:
<span class="citation">Matt Haynes, Alex Norton, Andrew McParland, and Rob Cooper; <cite>Speech-to-Text for Broadcasters, From Research to Implementation</cite>, SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018); SMPTE, 2018. Available at <a href="https://doi.org/10.5594/JMI.2018.2790658" target="_blank" rel="noopener">https://doi.org/10.5594/JMI.2018.2790658</a></span>
SMPTE's HTML Pub
Preview:
Matt Haynes, Alex Norton, Andrew McParland, and Rob Cooper; Speech-to-Text for Broadcasters, From Research to Implementation, SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018); SMPTE, 2018
doi: 10.5594/JMI.2018.2790658
url: https://doi.org/10.5594/JMI.2018.2790658
doi: 10.5594/JMI.2018.2790658
url: https://doi.org/10.5594/JMI.2018.2790658
Snippet:
<li> Matt Haynes, Alex Norton, Andrew McParland, and Rob Cooper; <cite id="bib-10-5594-jmi-2018-2790658">Speech-to-Text for Broadcasters, From Research to Implementation</cite>, SMPTE Motion Imaging Journal ( Volume: 127, Issue: 2, March 2018); SMPTE, 2018 <span class="doi">10.5594/JMI.2018.2790658</span> </li>