|Table of Contents|

Application of the Hidden Markov Model to the Classification of Passive Acoustic Signal(PDF)

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
1998年06期
Page:
481-485
Research Field:
Publishing date:

Info

Title:
Application of the Hidden Markov Model to the Classification of Passive Acoustic Signal
Author(s):
Ding Qinghai Zhuang Zhihong L u Jianwei Zhang Qingtai
School of Electronic Engineering and Optoelectronic Technology,NUST, Nanjing 2 10 0 94
Keywords:
acoust ic sig nal Marko v chains neural netw ork passive acoust ic signal classif icat io n Hidden Marko v Model
PACS:
TN911.7
DOI:
-
Abstract:
To improve the pro bability of passive acoust ic targ et ident ificat ion, the pr oblem of applicat ions of the hidden Marko v model ( HMM ) to the classif icatio n of passive acoust ic sig nal are first ly discussed in this paper. T hen , following the discussion abo ve , tw o combined classif iers are presented w hich are the mix ed featur e v ector HMM classif ier and hybrid HMM/ MLPNN classif ier . The result s show that the combined classif iers are superior to any individual specific classifier and have g reat po tentials in the field of passiv e acoust ic sig nal classif icatio n.

References:

1 Yang H, Amlan K . 2—D shape classificatio n using HMM . IEEE Tr ans o n PAMI, 1991,13( 11) : 1172~1184
2  Law rence R R. A tutor ial on HMM and selected applicatio n in speech r eco gnit ion .Proceeding s o f IEEE, 1989, 77( 2) : 257~286
3 Juang B H. The segment al K—Means algo rit hm for estimating par ameter s of HMM . IEEETrans on ASSP, 1989, 38( 7) : 1639~1641
4 包威权. 基于HMM/M LFNN 混合结构的说话人辨认研究. 北京大学学报( 自然科学版) ,1997, 33( 3) : 359~366

Memo

Memo:
-
Last Update: 2013-03-29