南開科技大學圖書館 |

Language: 繁體中文

說明(常見問題)

南開科技大學

圖書館首頁

編目中圖書申請

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Modulation scale analysis: Theory a...

Sukittanon, Somsak.

Modulation scale analysis: Theory and application for nonstationary signal classification.

紀錄類型:	書目-電子資源 : 單行本
正題名/作者:	Modulation scale analysis: Theory and application for nonstationary signal classification./
作者:	Sukittanon, Somsak.
面頁冊數:	132 p.
附註:	Source: Dissertation Abstracts International, Volume: 65-07, Section: B, page: 3621.
Contained By:	Dissertation Abstracts International65-07B.
標題:	Engineering, Electronics and Electrical. -
電子資源:	Download fulltext (下載全文)
ISBN:	0496866893

Modulation scale analysis: Theory and application for nonstationary signal classification.
Sukittanon, Somsak.

Modulation scale analysis: Theory and application for nonstationary signal classification. - 132 p.

Source: Dissertation Abstracts International, Volume: 65-07, Section: B, page: 3621.

Thesis (Ph.D.)--University of Washington, 2004.

The goal of this dissertation is to develop a new set of features for nonstationary signal classification (e.g. speech, music, or communication). Conventional features of nonstationary signals consist of series of short-term stationary features. These short-term features cannot sufficiently model the time-varying information of nonstationary signals without classifier memory and/or features with expanded temporal extent.

ISBN: 0496866893Subjects--Topical Terms:

170927
Engineering, Electronics and Electrical.

Modulation scale analysis: Theory and application for nonstationary signal classification.
LDR:02507nmm 2200289 4500 001 1000004305
005 20051102155244.5
008 141201s2004 eng d
020 $a 0496866893
035 $a (UnM)AAI3139548
035 $a AAI3139548
040 $a UnM $c UnM{me_controlnum}
100 1 $a Sukittanon, Somsak. $3 1000005627
245 1 0 $a Modulation scale analysis: Theory and application for nonstationary signal classification.
300 $a 132 p.
500 $a Source: Dissertation Abstracts International, Volume: 65-07, Section: B, page: 3621.
500 $a Chair: Les E. Atlas.
502 $a Thesis (Ph.D.)--University of Washington, 2004.
520 $a The goal of this dissertation is to develop a new set of features for nonstationary signal classification (e.g. speech, music, or communication). Conventional features of nonstationary signals consist of series of short-term stationary features. These short-term features cannot sufficiently model the time-varying information of nonstationary signals without classifier memory and/or features with expanded temporal extent.
520 $a To overcome deficiencies of short-term features, modulation scale features are proposed in this work. Partially motivated by psychoacoustics, the new features are derived from a wavelet transform of a time-frequency representation of signals. We discuss advantages of modulation scale over previous methods that used modulation spectra for classification, for example, the modulation spectrogram and cyclostationary analysis. The theory of modulation scale is provided along with an efficient implementation. Furthermore, we propose the derivation of channel compensation for modulation spectral features. The new method, called subband normalization, is not only robust to slowly varying convolutional noise, but also to time distortions (e.g. time-scale modification and time misalignment).
520 $a Finally, we assess the effectiveness of the new analysis on two applications: (1) content-based audio identification and (2) communication signal interception. A comparison between modulation scale features and other conventional short-term features is studied. The experiments are validated using both static and dynamic classifiers.
590 $a School code: 0250.
650 4 $a Engineering, Electronics and Electrical. $3 170927
690 $a 0544
710 2 0 $a University of Washington. $3 176174
773 0 $t Dissertation Abstracts International $g 65-07B.
790 1 0 $a Atlas, Les E., $e advisor
790 $a 0250
791 $a Ph.D.
792 $a 2004
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3139548 $z Download fulltext (下載全文)