Integrating Recent MLP Feature Extraction Techniques into TRAP Architecture

@inproceedings{Grzl2011IntegratingRM,
  title={Integrating Recent MLP Feature Extraction Techniques into TRAP Architecture},
  author={Frantisek Gr{\'e}zl and Martin Karafi{\'a}t},
  booktitle={INTERSPEECH},
  year={2011}
}
This paper is focused on the incorporation of recent techniques for multi-layer perceptron (MLP) based feature extraction in Temporal Pattern (TRAP) and Hidden Activation TRAP (HATS) feature extraction scheme. The TRAP scheme has been origin of various MLP-based features some of which are now indivisible part of state-of-the-art LVCSR systems. The modifications which brought most improvement – sub-phoneme targets and Bottle-Neck technique – are introduced into original TRAP scheme. Introduction… CONTINUE READING

From This Paper

Figures, tables, and topics from this paper.

References

Publications referenced by this paper.
Showing 1-10 of 11 references

Probabilistic and Bottle-Neck Features for LVCSR of Meetings

2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07 • 2007
View 3 Excerpts

TRAP-based probabilistic features for automatic speech recognition

F. Grézl
Ph.D. dissertation, Brno University of Technology, Czech Republic, 2007, http://www.fit.vutbr.cz/ grezl/publi/dis.pdf. • 2007
View 1 Excerpt

The AMI System for the Transcription of Speech in Meetings

2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07 • 2005
View 2 Excerpts

Similar Papers

Loading similar papers…