Failure simulation for a phoneme HMM based keyword spotter

A basic problem in keyword spotting is the fact that the keywords itself cannot be completely different from background speech. Therefore, false alarms arise from those parts of the keyword which are also contained in the background. The paper describes the favourable application of a model trellis which enables one to test individual phoneme sequences with respect to their influence on the underlying phoneme HMMs in a statistical way. It is shown, that the Viterbi path is highly affected by those partly fitting phoneme groups. The probability of occurrence of these phoneme sequences is captured by a statistical “speech model” consisting of a Markov graph having an order up to 2. In this way sequences of 1, 2, or 3 phonemes are considered. By combining the model trellis and the statistical speech model, the probability of false alarms can be precalculated in advance, thus providing an useful measure for the suitability of the keyword under consideration. When the choice of keywords was optimized by this suitability measure in a practical application (spotting multicom 94.4 data), the false alarm rate could be reduced by a factor of 3.5