國立虎尾科技大學 |

A Spectro-Temporal Framework for Compensation of Reverberation for Speech Recognition.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	A Spectro-Temporal Framework for Compensation of Reverberation for Speech Recognition./
作者:	Kumar, Kshitiz.
面頁冊數:	118 p.
附註:	Source: Dissertation Abstracts International, Volume: 72-08, Section: B, page: 4846.
Contained By:	Dissertation Abstracts International72-08B.
標題:	Engineering, Computer. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3459396
ISBN:	9781124702445

A Spectro-Temporal Framework for Compensation of Reverberation for Speech Recognition.
Kumar, Kshitiz.

A Spectro-Temporal Framework for Compensation of Reverberation for Speech Recognition. - 118 p.

Source: Dissertation Abstracts International, Volume: 72-08, Section: B, page: 4846.

Thesis (Ph.D.)--Carnegie Mellon University, 2011.

The objective of this thesis is the development of signal processing and analysis techniques that would provide sharply improved speech recognition accuracy in highly reverberant environments. Speech is a natural medium of communication for humans, and in the last decade various speech technologies like automatic speech recognition (ASR), voice response systems etc. have considerably matured. The above systems rely on the clarity of the captured speech but many of the real-world environments include noise and reverberation that mitigate the system performance. The key focus of the thesis is on the robustness of ASR to reverberation.

ISBN: 9781124702445Subjects--Topical Terms:

845407
Engineering, Computer.

A Spectro-Temporal Framework for Compensation of Reverberation for Speech Recognition.
LDR:03150nam 2200325 4500 001 712927
005 20121003100259.5
008 121101s2011 ||||||||||||||||| ||eng d
020 $a 9781124702445
035 $a (UMI)AAI3459396
035 $a AAI3459396
040 $a UMI $c UMI
100 1 $a Kumar, Kshitiz. $3 845405
245 1 2 $a A Spectro-Temporal Framework for Compensation of Reverberation for Speech Recognition.
300 $a 118 p.
500 $a Source: Dissertation Abstracts International, Volume: 72-08, Section: B, page: 4846.
500 $a Adviser: Richard Stern.
502 $a Thesis (Ph.D.)--Carnegie Mellon University, 2011.
520 $a The objective of this thesis is the development of signal processing and analysis techniques that would provide sharply improved speech recognition accuracy in highly reverberant environments. Speech is a natural medium of communication for humans, and in the last decade various speech technologies like automatic speech recognition (ASR), voice response systems etc. have considerably matured. The above systems rely on the clarity of the captured speech but many of the real-world environments include noise and reverberation that mitigate the system performance. The key focus of the thesis is on the robustness of ASR to reverberation.
520 $a In our work, we first provide a new framework to adequately and efficiently represent the problem of reverberation in speech feature domains. Although our framework incurs modeling approximation errors, we believe that it provides a good basis for developing reverberation compensation algorithms. Based on our framework, we successfully develop a number of dereverberation algorithms. The algorithms reduce the uncertainly involved in dereverberation tasks by using speech knowledge in terms of cepstral auto-correlation, cepstral distribution, and, non-negativity and sparsity of spectral values. We demonstrate the success of our algorithms on clean-training as well as matched-training.
520 $a Apart from dereverberation, we also provide an approach for noise robustness via a temporal-difference operation in the speech spectral domain. There, via a theoretical analysis, we predict an expected improvement in the SNR threshold shift for white-noise conditions. We also empirically quantify and study speech-feature level distortion with respect to speech-signal level additive noise.
520 $a Finally, we provide a new framework for a joint reverberation and noise representation and compensation. The new framework generalizes the spectral domain reverberation framework by incorporating an additive noise term. Working under the new framework, we combine our dereverberation and noise compensation approaches for better dereverberation as well as for the most challenging speech recognition task that includes both noise and reverberation components.
590 $a School code: 0041.
650 4 $a Engineering, Computer. $3 845407
650 4 $a Engineering, Electronics and Electrical. $3 845382
650 4 $a Computer Science. $3 593922
690 $a 0464
690 $a 0544
690 $a 0984
710 2 $a Carnegie Mellon University. $3 845406
773 0 $t Dissertation Abstracts International $g 72-08B.
790 1 0 $a Stern, Richard, $e advisor
790 $a 0041
791 $a Ph.D.
792 $a 2011
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3459396