High-quality Speech Coding with Sample RNN

Janusz Klejsa; Per Hedelin; Cong Zhou; Roy Fejgin; Lars F. Villemoes

ICASSP 2019accepted0 citations

High-quality Speech Coding with Sample RNN

Janusz Klejsa, Per Hedelin, Cong Zhou, Roy Fejgin, Lars F. Villemoes

Abstract

We provide a speech coding scheme employing a generative model based on SampleRNN that, while operating at significantly lower bitrates, matches or surpasses the perceptual quality of state-of-the-art classic wide-band codecs. Moreover, it is demonstrated that the proposed scheme can provide a meaningful rate-distortion trade-off without retraining. We evaluate the proposed scheme in a series of listening tests and discuss limitations of the approach.

BibTeX

@inproceedings{icassp2019_highqualityspeec,
  title = {High-quality Speech Coding with Sample RNN},
  author = {Janusz Klejsa and Per Hedelin and Cong Zhou and Roy Fejgin and Lars F. Villemoes},
  booktitle = {ICASSP 2019},
  year = {2019}
}