Combination of strongly and weakly constrained recognizers for reliable detection of OOVS
Lukáš Burget, Petr Schwarz, Pavel Matějka, Mirko Hannemann, Ariya Rastrow, CHRISTOPHER M. WHITE, Sanjeev P. Khudanpur, Hynek Heřmanský, Jaň Černocký · IEEE International Conference on Acoustics Speech and Signal Processing · 2008
This paper addresses the detection of OOV segments in the output of a large vocabulary continuous speech recognition (LVCSR) system. First, standard confidence measures from frame-based wordand phone- posteriors are investigated. Substantial improvement is obtained when posteriors from two systems — strongly constrained (LVCSR) and weakly constrained (phone posterior estimator) are combined. We show that this approach is also suitable for detection of general recognition errors. All results are presented on WSJ task with reduced recognition vocabulary.