A double-side WSOLA with gain prediction based on GRU for packet loss concealment

Patrick Mugisha, Jing Wang, Xiaohan Zhao, Zhuoran Li, Kai Qian

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

The perceived speech quality is one of the main issues in VoIP communications. Packet loss concealment (PLC) is among the best techniques to recover the lost packets over IP networks in order to improve the speech quality. In this paper, a double-side waveform similarity overlap and add(WSOLA) technique has been proposed in which both previously received packets and future packets involve in recovering the lost packets by using WSOLA technique. To overcome energy decrease that occurs during the waveform reconstruction, a gated recurrent unit (GRU) based gain control approach has been introduced to maintain the energy of a reconstructing waveform with the prediction of the lost packets gain by learning the previously received packets gain. The proposed algorithm has shown the outperformance to the standard WSOLA and double-side WSOLA without gain control at both random and burst packet loss according to the obtained perceptual evaluation of speech quality (PESQ) results.

Original languageEnglish
Title of host publicationICSIDP 2019 - IEEE International Conference on Signal, Information and Data Processing 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728123455
DOIs
Publication statusPublished - Dec 2019
Event2019 IEEE International Conference on Signal, Information and Data Processing, ICSIDP 2019 - Chongqing, China
Duration: 11 Dec 201913 Dec 2019

Publication series

NameICSIDP 2019 - IEEE International Conference on Signal, Information and Data Processing 2019

Conference

Conference2019 IEEE International Conference on Signal, Information and Data Processing, ICSIDP 2019
Country/TerritoryChina
CityChongqing
Period11/12/1913/12/19

Keywords

  • GRU
  • VoIP
  • WSOLA
  • gain prediction
  • packet loss concealment

Fingerprint

Dive into the research topics of 'A double-side WSOLA with gain prediction based on GRU for packet loss concealment'. Together they form a unique fingerprint.

Cite this