Parameter Estimation of a State-Space Model of Noise for Robust Speech Recognition

被引：3
|
作者：
Windmann, Stefan [1 ]
Haeb-Umbach, Reinhold [1 ]
机构：
[1] Univ Paderborn, Dept Commun Engn, D-33098 Paderborn, Germany
来源：
IEEE TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING | 2009年 / 17卷 / 08期
关键词：
Dynamical systems; noise estimation; robust speech recognition; FEATURE ENHANCEMENT; COMPENSATION;
D O I：
10.1109/TASL.2009.2023172
中图分类号：
O42 [声学];
学科分类号：
070206 ; 082403 ;
摘要：
In this paper, parameter estimation of a state-space model of noise or noisy speech cepstra is investigated. A blockwise EM algorithm is derived for the estimation of the state and observation noise covariance from noise-only input data. It is supposed to be used during the offline training mode of a speech recognizer. Further a sequential online EM algorithm is developed to adapt the observation noise covariance on noisy speech cepstra at its input. The estimated parameters are then used in model-based speech feature enhancement for noise-robust automatic speech recognition. Experiments on the AURORA4 database lead to improved recognition results with a linear state model compared to the assumption of stationary noise.
引用
收藏
页码：1577 / 1590
页数：14
相关论文
共 50 条
  • [1] Noise Moment and Parameter Estimation of State-Space Model
    Kost, Oliver
    Dunik, Jindrich
    Straka, Ondrej
    IFAC PAPERSONLINE, 2018, 51 (15): : 891 - 896
  • [2] State and parameter estimation of state-space model with entry-wise correlated uniform noise
    Pavelkova, Lenka
    Karny, Miroslav
    INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING, 2014, 28 (11) : 1189 - 1205
  • [3] Parameter estimation of exponentially damped sinusoids based on state-space model
    Yu X.-H.
    Zhang Z.-C.
    Li X.-B.
    Sun X.-D.
    Jilin Daxue Xuebao (Gongxueban)/Journal of Jilin University (Engineering and Technology Edition), 2019, 49 (06): : 2083 - 2088
  • [4] SPECTRAL ESTIMATION FOR NOISE ROBUST SPEECH RECOGNITION
    ERELL, A
    WEINTRAUB, M
    SPEECH AND NATURAL LANGUAGE, 1989, : 319 - 324
  • [5] Unsupervised noise model estimation for model-based robust speech recognition
    Graciarena, M
    Franco, H
    ASRU'03: 2003 IEEE WORKSHOP ON AUTOMATIC SPEECH RECOGNITION AND UNDERSTANDING ASRU '03, 2003, : 351 - 356
  • [6] Robust Estimation in Non-Linear State-Space Models With State-Dependent Noise
    Agamennoni, Gabriel
    Nebot, Eduardo M.
    IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2014, 62 (08) : 2165 - 2175
  • [7] Spectral Reconstruction and Noise Model Estimation Based on a Masking Model for Noise Robust Speech Recognition
    Gonzalez, Jose A.
    Gomez, Angel M.
    Peinado, Antonio M.
    Ma, Ning
    Barker, Jon
    CIRCUITS SYSTEMS AND SIGNAL PROCESSING, 2017, 36 (09) : 3731 - 3760
  • [8] Spectral Reconstruction and Noise Model Estimation Based on a Masking Model for Noise Robust Speech Recognition
    Jose A. Gonzalez
    Angel M. Gómez
    Antonio M. Peinado
    Ning Ma
    Jon Barker
    Circuits, Systems, and Signal Processing, 2017, 36 : 3731 - 3760
  • [9] Improved model parameter compensation methods for noise-robust speech recognition
    Chang, YH
    Chung, YJ
    Park, SU
    PROCEEDINGS OF THE 1998 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING, VOLS 1-6, 1998, : 561 - 564
  • [10] DYNAMIC SPEECH EMOTION RECOGNITION WITH STATE-SPACE MODELS
    Markov, Konstantin
    Matsui, Tomoko
    Septier, Francois
    Peters, Gareth
    2015 23RD EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO), 2015, : 2077 - 2081
12345