Speech recognition employing a set of Markov models that includes Markov models representing transitions to and from silence

摘要: Apparatus and method for constructing word baseforms which can be matched against a string of generated acoustic labels. A set phonetic phone machines are formed, wherein each machine has (i) plurality states, (ii) transitions extends from state to state, (iii) stored probability transition, (iv) label output probabilities, corresponding the producing label. The is formed include subset onset machines. probabilities macine correspond at least one element being uttered beginning speech segment. trailing single end Word constructed by concatenating selected set.