{"product_id":"ieej-rc13510-012","title":"制約を緩和した尤度計算を備えた単語HMMによる区間検出誤りに頑健な孤立単語認識","description":"\u003cp\u003e\u003cstrong\u003eカテゴリ: \u003c\/strong\u003e論文誌(論文単位)\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003eグループ名: \u003c\/strong\u003e【C】電子・情報・システム部門\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e発行日: \u003c\/strong\u003e2015\/10\/01\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003eタイトル(英語): \u003c\/strong\u003eRobust Speech Recognition against Misdetection Using Whole-Word HMMs and Relaxed Algorithm for Likelihood Calculation\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e著者名: \u003c\/strong\u003e早坂　昇（大阪電気通信大学情報通信工学部情報工学科）\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e著者名(英語): \u003c\/strong\u003eNoboru Hayasaka (Dept. of Engineering Informatics, Osaka Electro-Communication University)\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003eキーワード: \u003c\/strong\u003e孤立単語認識，区間検出誤り，単語モデル，制約緩和尤度計算　　Isolated-word recognition，Misdetection，Whole-word HMM，Relaxed algorithm for likelihood calculation\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e要約(英語): \u003c\/strong\u003eIn this paper, we propose a new framework that employs whole-word hidden Markov models with a relaxed algorithm for likelihood calculation. Although voice activity detection (VAD) is an essential technique in order to provide a sophisticated voice interface, a perfect VAD has not been proposed yet. An incomplete VAD leads to misdetections frequently. Since the conventional likelihood calculation, which has limitation in edge states, assumes that VAD works well, we cannot obtain sufficient recognition accuracy against the misdetections. The proposed method is a technique to improve the performance against misdetected speech in the framework which uses whole-word HMMs. The method employs an algorithm which relaxes the limitation of likelihood calculation in the edge states. In order to verify the effectiveness of the proposed method, we carried out recognition experiments for artificially shortened segments. With the conventional method, the average recognition rates were 91.81%, 91.50%, 89.56%, and 85.13% under 3, 4, 5, and 6 states\/phoneme conditions, respectively. However, the proposed method allowed us to achieve 91.88%, 92.64%, 92.74%, and 93.14% under the same conditions. Furthermore, we confirmed the effectiveness of the proposed method through experiments combined with CENSREC-1-C.\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e本誌: \u003c\/strong\u003e\u003ca href=\"\/products\/ieej-rc13510\"\u003e電気学会論文誌C（電子・情報・システム部門誌） Vol.135 No.10 （2015） 特集：産業界を支える情報通信技術\u003c\/a\u003e\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e本誌掲載ページ: \u003c\/strong\u003e1236-1243 p\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e原稿種別: \u003c\/strong\u003e論文／日本語\u003c\/p\u003e\u003cp\u003e\u003cstrong\u003e電子版へのリンク: \u003c\/strong\u003e\u003ca target=\"_blank\" href=\"https:\/\/www.jstage.jst.go.jp\/article\/ieejeiss\/135\/10\/135_1236\/_article\/-char\/ja\/\"\u003ehttps:\/\/www.jstage.jst.go.jp\/article\/ieejeiss\/135\/10\/135_1236\/_article\/-char\/ja\/\u003c\/a\u003e\u003c\/p\u003e","brand":"IEEJ-P10","offers":[{"title":"冊子印刷（一般価格770円\/会員価格550円） \/ A4 \/ 8","offer_id":46349472760047,"sku":"IEEJ-RC13510-012-PRT","price":770.0,"currency_code":"JPY","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0718\/9512\/2159\/files\/IEEJ-RC13510_4ff82815-686b-4ba9-ab21-82cfa571f067.png?v=1747382579","url":"https:\/\/ieej.bookpark.ne.jp\/products\/ieej-rc13510-012","provider":"電気学会 電子図書館","version":"1.0","type":"link"}