ISCA Archive Eurospeech 1997
ISCA Archive Eurospeech 1997

Model-based approach for robust speech recognition in noisy environements with multiple noise sources

Do Yeong Kim, Nam Soo Kim, Chong Kwan Un

In this paper, we consider the hidden Markov model(HMM) parameter compensation in noisy environments with multiple noise sources based on the vector Taylor series (VTS) approach. General formulations for multiple environmental variables are derived and systematic expectation-maximization (EM) solutions are presented in maximum likelihood (ML) sense. It is assumed that each noise source is independent and having Gaussian distribution. To evaluate proposed method, we conduct speaker independent isolated word recognition experiments in various noisy environments. Experimental results show that proposed algorithm ahieves significant improvement. Especially, the proposed method is consistently more effective than the parallel model combination (PMC) based on log-normal approximation.


doi: 10.21437/Eurospeech.1997-115

Cite as: Kim, D.Y., Kim, N.S., Un, C.K. (1997) Model-based approach for robust speech recognition in noisy environements with multiple noise sources. Proc. 5th European Conference on Speech Communication and Technology (Eurospeech 1997), 1123-1126, doi: 10.21437/Eurospeech.1997-115

@inproceedings{kim97b_eurospeech,
  author={Do Yeong Kim and Nam Soo Kim and Chong Kwan Un},
  title={{Model-based approach for robust speech recognition in noisy environements with multiple noise sources}},
  year=1997,
  booktitle={Proc. 5th European Conference on Speech Communication and Technology (Eurospeech 1997)},
  pages={1123--1126},
  doi={10.21437/Eurospeech.1997-115},
  issn={1018-4074}
}