ESCA-NATO Robust Workshop '97: Robust Error Correction
of Continuous Speech Recognition
Please download a copy of the paper by clicking here:
ESCA-NATO Robust Workshop '97 Paper in gzip'ed PostScript format
(~28K).
Abstract:
We present a post-processing technique for correcting errors committed
by an arbitrary continuous speech recognizer. The technique leverages
our observation that consistent recognition errors arising from
mismatched training and usage conditions can be modeled and corrected.
We have implemented a post-processor called SPEECHPP to correct
word-level errors, and we show that this post-processing technique
applies successfully when the training and usage domains differ even
slightly; for the purposes of the recognizer, such a difference
manifests itself as differences in the vocabulary and in the
likelihoods of word collocations. We hypothesize that other
differences between the training and usage conditions yield
recognition errors with some consistency also. Hence, we propose that
our technique be used to compensate for those mismatches as well.
Last Change 97 January 30 by
ringger@cs.rochester.edu