Title :
An adaptive-beam pruning technique for continuous speech recognition
Author :
Hamme, Hugo Van ; Aelten, Filip Van
Author_Institution :
Lernout & Hauspie Speech Products NV, Wemmel, Belgium
Abstract :
Pruning is an essential paradigm to build HMM based large vocabulary speech recognisers that use reasonable computing resources. Unlikely sentence, word or subword hypotheses are removed from the search space when their likelihood falls outside a beam relative to the best scoring hypothesis. A method for automatically steering this beam such that the search space attains a predefined size is presented
Keywords :
adaptive systems; hidden Markov models; search problems; speech processing; speech recognition; HMM based large vocabulary speech recognisers; adaptive beam pruning technique; computing resources; continuous speech recognition; likelihood; predefined size; scoring hypothesis; search space; subword hypotheses; Acoustic beams; Automatic control; Automatic speech recognition; Data mining; Decoding; Hidden Markov models; Histograms; Size control; Speech recognition; Vocabulary;
Conference_Titel :
Spoken Language, 1996. ICSLP 96. Proceedings., Fourth International Conference on
Conference_Location :
Philadelphia, PA
Print_ISBN :
0-7803-3555-4
DOI :
10.1109/ICSLP.1996.607212
