Rejection techniques for digit recognition in telecommunication applications

L. Villarrubia, A. Acero
1993 IEEE International Conference on Acoustics Speech and Signal Processing  
In this paper we describe a technique for non-keyword rejection and we will evaluate in the context of a n audiotex service using the ten Spanish digits. The baseline keyword recognition system is a speaker-independent continuous density Hidden Markov Model recognizer. We propose the use of an affine transformation to the log-probability of the garbage model, a n HMM model trained to account for both nonkeyword speech and non-stationary telephone noises. The parameters of the transformation for
more » ... the case of isolated keywords are chosen to minimize a cost function that weighs the keyword error rate, keyword rejection rate and false acceptance rate according to the a priori probabilities of keywordhon-keyword and the requirements of the specific application. This technique was also extended to embedded keywords (word-spotting). Use of this rejection technique on the audiotex application reduced the total cost function up to 20% for isolated-word case and 12% for the word-spotting case. 11- 455 0-7803-0946-4/93 $3.00 ' 1993 IEEE Authorized licensed use limited to: MICROSOFT. Downloaded on December 24, 2008 at 03:45 from IEEE Xplore. Restrictions apply.
doi:10.1109/icassp.1993.319339 dblp:conf/icassp/VillarrubiaA93 fatcat:2dxeruq4nbgv3ae6yzfpms6gdu