頻域獨立成分分析法於語音訊號分離之研究

Full metadata record

DC Field	Value	Language
dc.contributor.author	連憶如	en_US
dc.contributor.author	Yi-Ru Lian	en_US
dc.contributor.author	胡竹生	en_US
dc.contributor.author	Jwu-Shen Hu	en_US
dc.date.accessioned	2014-12-12T01:38:26Z	-
dc.date.available	2014-12-12T01:38:26Z	-
dc.date.issued	2003	en_US
dc.identifier.uri	http://140.113.39.130/cdrfb3/record/nctu/#GT009112508	en_US
dc.identifier.uri	http://hdl.handle.net/11536/44591	-
dc.description.abstract	在本論文中，針對由聲源與channel捲積(convolute)而來之混和語音訊號在頻域利用獨立成分分析法做訊號分離。將對於頻域獨立成分分析法所延伸出之兩大問題- permutation problem與dilation problem加以探討。由於混和訊號需滿足複數之FastICA演算法對於其訊號源之假設，而一般語音訊號無法符合其假設，因此設計出一組適合FastICA分離之訊號源使得FastICA能夠成功地完成訊號分離以找出其對應之demixing matrix，而達成訊號分離之目的。除此之外，當訊號源為一固定噪音與一移動語音時，亦可利用noise之channel效應不變之特性將語音分離出來。	zh_TW
dc.language.iso	zh_TW	en_US
dc.subject	獨立成分分析	zh_TW
dc.subject	語音訊號	zh_TW
dc.subject	訊號分離	zh_TW
dc.subject	ICA	en_US
dc.subject	speech signal	en_US
dc.subject	signal separation	en_US
dc.title	頻域獨立成分分析法於語音訊號分離之研究	zh_TW
dc.title	An investigation of frequency domain ICA for speech signal separation	en_US
dc.type	Thesis	en_US
dc.contributor.department	電控工程研究所	zh_TW
Appears in Collections:	Thesis

Files in This Item:

250801.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.