Baseline Wsj Acoustic Models for Htk and Sphinx : Training Recipes and Recognition Experiments

@inproceedings{Vertanen2007BaselineWA,
  title={Baseline Wsj Acoustic Models for Htk and Sphinx : Training Recipes and Recognition Experiments},
  author={Keith Vertanen},
  year={2007}
}
For speech recognition research, it is often necessary to start with a competent baseline acoustic model. But training and tuning a competent model using research recognizers such as Cambridge’s HTK and CMU’s Sphinx can be time-consuming. In an effort to minimize wasted effort, I have created recipes for HTK and Sphinx which utilize the standard Wall Street Journal training corpus. In this paper, these recipes are described. The word error rate (WER) and real-time performance of the models are… CONTINUE READING
Highly Cited
This paper has 136 citations. REVIEW CITATIONS

Citations

Publications citing this paper.
Showing 1-10 of 97 extracted citations

136 Citations

0102030'09'11'13'15'17
Citations per Year
Semantic Scholar estimates that this publication has 136 citations based on the available data.

See our FAQ for additional information.

References

Publications referenced by this paper.
Showing 1-4 of 4 references

CMU Sphinx Wall Street Journal Training Recipe,

  • Keith Vertanen
  • http://www.inference.phy. cam.ac.uk/kv227/sphinx…
  • 2006
1 Excerpt

HTK Wall Street Journal Training Recipe,

  • Keith Vertanen
  • http://www.inference.phy. cam.ac.uk/kv227/htk/,
  • 2006
1 Excerpt

and P

  • S. Young, D. Ollason, V. Valtchev
  • Woodland, The HTK Book (for HTK Version 3.2…
  • 2002
2 Excerpts

Similar Papers

Loading similar papers…