Full metadata record
DC FieldValueLanguage
dc.contributor.author楊智合en_US
dc.contributor.author江振宇en_US
dc.contributor.author劉銘傑en_US
dc.contributor.author王逸如en_US
dc.contributor.author廖元甫en_US
dc.contributor.author陳信宏en_US
dc.date.accessioned2014-12-16T06:12:05Z-
dc.date.available2014-12-16T06:12:05Z-
dc.date.issued2014-06-11en_US
dc.identifier.govdocG10L015/00zh_TW
dc.identifier.urihttp://hdl.handle.net/11536/103398-
dc.description.abstract本發明係揭露一種中文語音辨識裝置及其辨識方法,首先,接收一語音訊號,以進行辨識後,輸出一格狀詞組(word lattice)。接著,接收格狀詞組,並根據一韻律停頓模型、一韻律狀態模型、一音節韻律聲學模型、一音節間韻律聲學模型與一因子化語言模型,重新計算格狀詞組中詞弧上的分數,將其重新排名,以輸出語音訊號對應之一語言標籤、一韻律標籤與一音段標記。本發明利用兩階段方式重新計分,不僅可提升基本語音資訊的辨識率,更可標記出語言、韻律、音段等標籤,以供後級語音技術所需的韻律結構與語言資訊。zh_TW
dc.language.isozh_TWen_US
dc.title中文語音辨識裝置及其辨識方法zh_TW
dc.typePatentsen_US
dc.citation.patentcountryTWNzh_TW
dc.citation.patentnumberI441163zh_TW
Appears in Collections:Patents


Files in This Item:

  1. I441163.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.