7th International Conference on Spoken Language Processing

September 16-20, 2002
Denver, Colorado, USA

Speech Enhancement Using Wavelet Packet Transform

Sungwook Chang, Sungil Jung, Y. Kwon, Sung-il Yang

Hanyang University, Korea

In this paper, we present denoising algorithm using wavelet packet transform based on VAD(Voice Activity Detection). For the purpose, the weighting parameter estimation method based on speech dominant indicator and the modified Beroutiís method are suggested. Subjective evaluation shows that the performance of the proposed algorithm with wavelet bases is better than that with Fourier basis. Log-Domain Speech Feature Enhancement Using

Full Paper

Bibliographic reference.  Chang, Sungwook / Jung, Sungil / Kwon, Y. / Yang, Sung-il (2002): "Speech enhancement using wavelet packet transform", In ICSLP-2002, 1809-1812.