keywords hidden markov models; sensors; accuracy; training; speech recognition; data models; databases; speech recognition; hidden markov models; learning (artificial intelligence); maximum likelihood estimation