Fabricating conversational speech data with acoustic models: a program to examine model-data mismatch

@inproceedings{McAllaster1998FabricatingCS,
  title={Fabricating conversational speech data with acoustic models: a program to examine model-data mismatch},
  author={Don McAllaster and Larry Gillick and Francesco Scattone and Michael Newman},
  booktitle={ICSLP},
  year={1998}
}
We present a study of data simulated using acoustic models trained on Switchboard data, and then recognized using various Switchboard-trained acoustic models. When we recognize real Switchboard conversations, simple development models give a word error rate (WER) of about 47 percent. If instead we simulate the speech data using word transcriptions of the conversation, obtaining the pronunciations for the words from our recognition dictionary, the WER drops by a factor of five to ten. In a third… CONTINUE READING
Highly Cited
This paper has 88 citations. REVIEW CITATIONS

Citations

Publications citing this paper.
Showing 1-10 of 58 extracted citations

An automatic and adaptive phonetic transcriber for the Basque language

2015 4th International Work Conference on Bioinspired Intelligence (IWOBI) • 2015
View 1 Excerpt

89 Citations

0510'99'03'08'13'18
Citations per Year
Semantic Scholar estimates that this publication has 89 citations based on the available data.

See our FAQ for additional information.

References

Publications referenced by this paper.
Showing 1-2 of 2 references

Pronunciation Modeling for Conversational Speech Recognition: A Status Report from WS97,

B. Byrne
IEEE ASRU Workshop, • 1997
View 1 Excerpt

Similar Papers

Loading similar papers…