ISCA Archive Interspeech 2008
ISCA Archive Interspeech 2008

In-car speech recognition using model-based wiener filter and multi-condition training

Masanori Tsujikawa, Takayuki Arakawa, Ryosuke Isotani

This paper presents in-car speech recognition using a model-based Wiener filter (MBW) and multi-condition (MC) training. The MBW is a 2-step denoising algorithm based on both rough and precise estimation of speech signals. Correcting roughly estimated signals with a Gaussian mixture model (GMM) makes it possible to accurately denoise with little computational cost. In an evaluation of in-car speech recognition, training of both a GMM and a back-end hidden Markov model (HMM) was performed using both studio-recorded speech signals as well as those signals mixed with in-car noise signals that were recorded in real car environments. In-car speech signals for testing were recorded with a plurality of microphones in different car environments. With respect to word accuracy obtained with MC-trained HMM, it was confirmed that the MBW with MC-trained GMM outperformed the Noise Reduction in ETSI advanced front-end.


doi: 10.21437/Interspeech.2008-284

Cite as: Tsujikawa, M., Arakawa, T., Isotani, R. (2008) In-car speech recognition using model-based wiener filter and multi-condition training. Proc. Interspeech 2008, 972-975, doi: 10.21437/Interspeech.2008-284

@inproceedings{tsujikawa08_interspeech,
  author={Masanori Tsujikawa and Takayuki Arakawa and Ryosuke Isotani},
  title={{In-car speech recognition using model-based wiener filter and multi-condition training}},
  year=2008,
  booktitle={Proc. Interspeech 2008},
  pages={972--975},
  doi={10.21437/Interspeech.2008-284}
}