Temporal patterns (TRAPs) in ASR of noisy speech

被引：82

作者：

Hermansky, H ^{[1
]}

Sharma, S ^{[1
]}

机构：

[1] Oregon Grad Inst Sci & Technol, Portland, OR 97208 USA

来源：

ICASSP '99: 1999 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, PROCEEDINGS VOLS I-VI | 1999年

关键词：

D O I：

10.1109/ICASSP.1999.758119

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

In this paper we study a new approach to processing temporal information for automatic speech recognition (ASR). Specifically, we study the use of rather longtime TempRAl Patterns (TRAPs) of spectral energies in place of the conventional spectral patterns for ASR. The proposed Neural TRAPs are found to yield significant amount of complementary information to that of the conventional spectral feature based ASR system. A combination of these two ASR systems is shown to result in improved robustness to several types of additive and convolutive environmental degradations.

引用

页码：289 / 292

页数：4

共 11 条

[1] How Do Humans Process and Recognize Speech? [J].

Allen, Jont B. .

IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, 1994, 2 (04) :567-577

[2]

BOURLARD H, 1996, P ICSLP 96, V1, P426

[3]

Bourlard H. A., 1994, Connectionist speech recognition: a hybrid approach

[4]

COLE R, 1994, P ICSLP 94 SEPT

[5]

COLE RA, 1995, P EUR SEPT, V95, P821

[6] PERCEPTUAL LINEAR PREDICTIVE (PLP) ANALYSIS OF SPEECH [J].

HERMANSKY, H .

JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA, 1990, 87 (04) :1738-1752

[7] Should recognizers have ears? [J].

Hermansky, H .

SPEECH COMMUNICATION, 1998, 25 (1-3) :3-27

[8]

HERMANSKY H, 1998, IN PRESS P ICSLP 98

[9]

SHARMA S, 1998, P SPEAK REC ITS COMM

[10]

TIBREWALA S, 1997, P IEEE INT C AC SPEE, V2, P1255

← 1 2 →