next up previous
次へ: Experimental conditions for sentence 上へ: Experiments for Sentence Speech 戻る: Experiments for Sentence Speech

Speech recognition

To show the effectiveness of an Ergodic HMM as a language model, we carried out a sentence speech recognition experiment. A standard speech recognition algorithm employed a continuous mixture HMM and beam search, with an Ergodic HMM providing the language information.



Jin'ichi Murakami 平成13年1月19日