ISCA Archive ISCSLP 2002
ISCA Archive ISCSLP 2002

An investigation on wireless speech recognition by data contamination and robust training techniques

Wei-Tyng Hong, Ke-Shiu Chen

This paper is concerned with the robust endpoint detection and noisy speech recognition over wireless network. Firstly, the MLP-based and GMM-based endpoint detection incorporated with data contamination and continuous spectral subtraction techniques were investigated. Then, for noisy wireless speech recognition, a combined technique of data contamination and robust training was proposed to separately model the environmental characteristics and phonetic information. According to the results from an abbreviated stock name recognition task, we observe that the proposed techniques has the potential to improve robustness not only on diverse data contaminated training data, but also on the unmatched noise-type condition between training and testing environments.


Cite as: Hong, W.-T., Chen, K.-S. (2002) An investigation on wireless speech recognition by data contamination and robust training techniques. Proc. International Symposium on Chinese Spoken Language Processing, paper 51

@inproceedings{hong02_iscslp,
  author={Wei-Tyng Hong and Ke-Shiu Chen},
  title={{An investigation on wireless speech recognition by data contamination and robust training techniques}},
  year=2002,
  booktitle={Proc. International Symposium on Chinese Spoken Language Processing},
  pages={paper 51}
}