東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Linked to FindBook

Google Book

Amazon

博客來

Computational spectrotemporal auditory model with applications to acoustical information processing.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Computational spectrotemporal auditory model with applications to acoustical information processing./
Author:	Chi, Tai-Shih.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2003,
Description:	137 p.
Notes:	Source: Dissertations Abstracts International, Volume: 65-05, Section: B.
Contained By:	Dissertations Abstracts International65-05B.
Subject:	Electrical engineering. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3094467
ISBN:	9780496421688

Computational spectrotemporal auditory model with applications to acoustical information processing.
Chi, Tai-Shih.

Computational spectrotemporal auditory model with applications to acoustical information processing. - Ann Arbor : ProQuest Dissertations & Theses, 2003 - 137 p.

Source: Dissertations Abstracts International, Volume: 65-05, Section: B.

Thesis (Ph.D.)--University of Maryland, College Park, 2003.

This item must not be sold to any third party vendors.

A computational spectrotemporal auditory model based on neurophysiological findings in early auditory and cortical stages is described. The model provides a unified multiresolution representation of the spectral and temporal features of sound likely critical in the perception of timbre. Several types of complex stimuli are used to demonstrate the spectrotemporal information preserved by the model. Shown by these examples, this two stage model reflects the apparent progressive loss of temporal dynamics along the auditory pathway from the rapid phase-locking (several kHz in auditory nerve), to moderate rates of synchrony (several hundred Hz in midbrain), to much lower rates of modulations in the cortex (around 30 Hz). To complete this model, several projection-based reconstruction algorithms are implemented to resynthesize the sound from the representations with reduced dynamics. One particular application of this model is to assess speech intelligibility. The spectro-temporal Modulation Transfer Functions (MTF) of this model is investigated and shown to be consistent with the salient trends in the human MTFs (derived from human detection thresholds) which exhibit a lowpass function with respect to both spectral and temporal dimensions, with 50% bandwidths of about 16 Hz and 2 cycles/octave. Therefore, the model is used to demonstrate the potential relevance of these MTFs to the assessment of speech intelligibility in noise and reverberant conditions. Another useful feature is the phase singularity emerged in the scale space generated by this multiscale auditory model. The singularity is shown to have certain robust properties and carry the crucial information about the spectral profile. Such claim is justified by perceptually tolerable resynthesized sounds from the nonconvex singularity set. In addition, the singularity set is demonstrated to encode the pitch and formants at different scales. These properties make the singularity set very suitable for traditional speech tasks such as vowel recognition or speaker (music instrument) identification. Other potential applications and future modification of this model are also discussed at the end.

ISBN: 9780496421688Subjects--Topical Terms:

649834
Electrical engineering.
Subjects--Index Terms:

Acoustical

Computational spectrotemporal auditory model with applications to acoustical information processing.
LDR:03448nmm a2200373 4500 001 2348677
005 20220912135632.5
008 241004s2003 ||||||||||||||||| ||eng d
020 $a 9780496421688
035 $a (MiAaPQ)AAI3094467
035 $a AAI3094467
040 $a MiAaPQ $c MiAaPQ
100 1 $a Chi, Tai-Shih. $3 3688046
245 1 0 $a Computational spectrotemporal auditory model with applications to acoustical information processing.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2003
300 $a 137 p.
500 $a Source: Dissertations Abstracts International, Volume: 65-05, Section: B.
500 $a Publisher info.: Dissertation/Thesis.
500 $a Advisor: Shamma, Shihab A.
502 $a Thesis (Ph.D.)--University of Maryland, College Park, 2003.
506 $a This item must not be sold to any third party vendors.
506 $a This item must not be added to any third party search indexes.
520 $a A computational spectrotemporal auditory model based on neurophysiological findings in early auditory and cortical stages is described. The model provides a unified multiresolution representation of the spectral and temporal features of sound likely critical in the perception of timbre. Several types of complex stimuli are used to demonstrate the spectrotemporal information preserved by the model. Shown by these examples, this two stage model reflects the apparent progressive loss of temporal dynamics along the auditory pathway from the rapid phase-locking (several kHz in auditory nerve), to moderate rates of synchrony (several hundred Hz in midbrain), to much lower rates of modulations in the cortex (around 30 Hz). To complete this model, several projection-based reconstruction algorithms are implemented to resynthesize the sound from the representations with reduced dynamics. One particular application of this model is to assess speech intelligibility. The spectro-temporal Modulation Transfer Functions (MTF) of this model is investigated and shown to be consistent with the salient trends in the human MTFs (derived from human detection thresholds) which exhibit a lowpass function with respect to both spectral and temporal dimensions, with 50% bandwidths of about 16 Hz and 2 cycles/octave. Therefore, the model is used to demonstrate the potential relevance of these MTFs to the assessment of speech intelligibility in noise and reverberant conditions. Another useful feature is the phase singularity emerged in the scale space generated by this multiscale auditory model. The singularity is shown to have certain robust properties and carry the crucial information about the spectral profile. Such claim is justified by perceptually tolerable resynthesized sounds from the nonconvex singularity set. In addition, the singularity set is demonstrated to encode the pitch and formants at different scales. These properties make the singularity set very suitable for traditional speech tasks such as vowel recognition or speaker (music instrument) identification. Other potential applications and future modification of this model are also discussed at the end.
590 $a School code: 0117.
650 4 $a Electrical engineering. $3 649834
650 4 $a Acoustics. $3 879105
653 $a Acoustical
653 $a Auditory
653 $a Information processing
653 $a Spectrotemporal
690 $a 0544
690 $a 0986
710 2 $a University of Maryland, College Park. $3 657686
773 0 $t Dissertations Abstracts International $g 65-05B.
790 $a 0117
791 $a Ph.D.
792 $a 2003
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3094467