汽車多媒體應用可個人化之強健語音辨識系統;A Robust and Personalizable Speech Recognition System for Multimedia Applications in a Car

NCUIR > College of Electrical Engineering & Computer Science > Department of Electrical Engineering > Research Project > Item 987654321/57420

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/57420

Title:	汽車多媒體應用可個人化之強健語音辨識系統;A Robust and Personalizable Speech Recognition System for Multimedia Applications in a Car
Authors:	莊堯棠
Contributors:	中央大學電機工程學系
Keywords:	資訊科學--軟體
Date:	2009-09-01
Issue Date:	2012-10-01 15:20:24 (UTC+8)
Publisher:	行政院國家科學委員會
Abstract:	隨著車載電子多媒體系統的迅速發展，使得車內提供的服務無限可能，而語音辨識將扮演其中一個關鍵性的角色。此計畫模擬在汽車環境之下使用多媒體系統的情況，將發展一套具強健性且兼有個人化功能的多媒體應用之語音辨識系統。在上年度的計畫中（97/08～98/07），我們所研製完成的系統已具有：以語音來選播音樂、撥打電話、上網瀏覽及尋求導航系統等四大項服務的功能。並以語音合成的方式模擬人聲，作為系統詢問和回答的方式，讓使用者感到舒適友善。延續上年度成果，本計畫擬定為一兩年期的計畫。第一年裡（98/08～ 99/07），我們計畫處理汽車在雜訊環境之下的語音辨識問題，研究汽車內強健性語音辨識的方法，我們擬運用強健語音參數、模型調適等技術，或研發其他方法來降低測試環境對語音辨識的影響。在第二年的計畫中（99/08～100/07），我們將進一步對系統增建個人化使用的設計，先識別或確認語者，再運用模型調適技術，發展出以使用者語音為主的辨識系統，並強化系統的整體效能。 ; A vehicle electronic multimedia system makes the car services friendly. In the system, a reliable speech recognition system will play a very important role. In this project, we will develop a robust and personalizable speech recognition system for a multimedia application in a car environment. In the last project (08/2008~07/2009), we have implemented a speech recognition system which is able to provide four services, i.e. listening to the selected music, making a phone call, using the internet, and inquiring for a navigation system. The user-friendly interface is made through an interactive dialogue approach. Speech synthesis is adopted in the system to simulate human voices as response. Following the above-mentioned project, this is a two-year project. In the first year (08/2009~07/2010), we are going to deal with the robustness problem of speech recognition against noise in a car environment. Robust features and model adaptation as well as other methods will be considered to reduce the environmental impact. In the 2nd year of this project (08/2010~07/2011), the recognition system will be further strengthened. By using speaker recognition/verification techniques to identify the user and then adapting the speech models for the user, the performance of the personalized recognition system is expected to be improved further. ; 研究期間 9808 ~ 9907
Relation:	財團法人國家實驗研究院科技政策研究與資訊中心
Appears in Collections:	[Department of Electrical Engineering] Research Project

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	296	View/Open

社群 sharing

Loading...