Keyword Spotting Using Normalization of Posterior Probability Confidence Measures

Keyword Spotting Using Normalization of Posterior Probability Confidence Measures by Rachna Vijay Vargiya Thesis Advisor: Marius C. Silaghi, Ph.D. Keyword spotting techniques deal with recognition of predefined vocabulary keywords from a voice stream. This research uses HMM based keyword spotting algorithms for this purpose. The three most important componenets of a keyword detection system are confidence measure, pruning technique and evaluation of results. We suggest that best match for a keyword would be an alignment in which all constituent states have high emission probabilities. Therefore score of even the worst subsequence must also be better than a thershold and a path can be represented by the score of its worst subsequence match. This confidence measure is called Real Fitting. The harsher the pruning in a technique, the fewer paths survive. This increases the speed as well as the risk of pruning the best match. Three levels of pruning are explored and results and performance are compared. Since the proposed algorithms do no follow