Please use this identifier to cite or link to this item: https://doi.org/10.1109/ICASSP.2012.6288983
Title: Implicit trajectory modelling using temporally varying weight regression for automatic speech recognition
Authors: Liu, S.
Sim, K.C. 
Keywords: complexity control
nonlinear constrained optimization
regression
trajectory modelling
Issue Date: 2012
Source: Liu, S.,Sim, K.C. (2012). Implicit trajectory modelling using temporally varying weight regression for automatic speech recognition. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings : 4761-4764. ScholarBank@NUS Repository. https://doi.org/10.1109/ICASSP.2012.6288983
Abstract: Recently, implicit trajectory modelling using temporally varying model parameters has achieved promising gains over the discriminatively trained standard HMM system. However, these works only focus on the temporally varying means or precisions explicitly. It is interesting to explore the capability of temporally varying weights, since the effect of time varying Gaussian parameters can be achieved by adjusting the weights of Gaussian Mixture Models (GMM) for different observation. This paper proposes a Temporally Varying Weight Regression (TVWR) model to learn the importance of different Gaussian components under different temporal contexts. Technically, TVWR factorizes the HMM state likelihood such that the contextual information can be modelled using time varying weights. Additionally, approximate constraints are derived to ensure a valid probabilistic model for TVWR. Experimental results for continuous speech recognition on Wall Street Journal show consistent improvements with varying system complexity and about 12% relative significant improvements in the best case. © 2012 IEEE.
Source Title: ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
URI: http://scholarbank.nus.edu.sg/handle/10635/39896
ISBN: 9781467300469
ISSN: 15206149
DOI: 10.1109/ICASSP.2012.6288983
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

6
checked on Dec 13, 2017

Page view(s)

40
checked on Dec 9, 2017

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.