Login / Signup

A Preliminary Study of Employing Lowpass-Filtered and Time-Reversed Feature Sequences as Data Augmentation for Speech Enhancement Deep Networks.

Che-Wei LiaoPing-Chen WuJeih-weih Hung
Published in: ISPACS (2022)
Keyphrases
  • low pass
  • hidden markov models
  • high quality
  • discrete wavelet transform
  • feature selection