Segment generation and clustering in the HTK broadcast news transcription system

@inproceedings{Hain1998SegmentGA,
  title={Segment generation and clustering in the HTK broadcast news transcription system},
  author={Thomas Hain and Sue E. Johnson and Andreas Tuerk and Philip C. Woodland and Steve J. Young},
  year={1998}
}
This paper describes the segmentation, gender detection and segment clustering scheme used in the 1997 HTK broadcast news evaluation system and presents results on both the unpartitioned 1996 development and the 1997 evaluation sets. The stages of our approach are presented, namely classification, segmentation and gender detection, gender relabelling, and clustering of speech segments. The evaluation audio stream has been segmented according to audio type with a frame accuracy up to 95… CONTINUE READING

Figures, Tables, Results, and Topics from this paper.

Key Quantitative Results

  • The evaluation audio stream has been segmented according to audio type with a frame accuracy up to 95%.
  • The evaluation audio stream has been segmented according toaudio type with a frame accuracy up to 95%.

Citations

Publications citing this paper.
SHOWING 1-10 OF 78 CITATIONS

Efficient audio stream segmentation via the combined T/sup 2/ statistic and Bayesian information criterion

  • IEEE Transactions on Speech and Audio Processing
  • 2005
VIEW 4 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Multiple events tracking in sound tracks

  • 2004 IEEE International Conference on Multimedia and Expo (ICME) (IEEE Cat. No.04TH8763)
  • 2004
VIEW 3 EXCERPTS
CITES METHODS
HIGHLY INFLUENCED

Improved DNN-based segmentation for multi-genre broadcast audio

  • 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • 2016

Model-based speech/non-speech segmentation of a heterogeneous multilingual TV broadcast collection

  • 2013 International Symposium on Intelligent Signal Processing and Communication Systems
  • 2013
VIEW 1 EXCERPT
CITES METHODS

FILTER CITATIONS BY YEAR

1997
2016

CITATION STATISTICS

  • 3 Highly Influenced Citations

References

Publications referenced by this paper.
SHOWING 1-7 OF 7 REFERENCES

DARPA Broadcast News Transcription and Understanding Workshop, Virginia

P. C. Woodland, T. Hain, +3 authors S. J. Young
  • HTK Broadcast News Transcription System. to appear in Proc
  • 1998
VIEW 3 EXCERPTS
HIGHLY INFLUENTIAL

Maximum Likelihoo d Linear Regression for Speaker Adaptation of Continuous Den sity Hidden Markov Models.Computer

C. J. Leggetter
  • Speech & Language,
  • 1995
VIEW 7 EXCERPTS
HIGHLY INFLUENTIAL