Abstract
This paper describes our work on real-time two-party and multi-party VoIP (voice-over-IP) systems that can achieve high perceptual conversational quality. It focuses on the fundamental understanding of conversational quality and its trade-offs among the design of speech codecs and strategies for network control, playout scheduling, and loss concealments. We have studied three key aspects that address the limitations of existing work and improve the perceptual quality of VoIP systems. Firstly, we have developed a statistical approach based on just-noticeable difference (JND) to significantly reduce the large number of subjective tests, as well as a classification method to automatically learn and generalize the results to unseen conditions. Using network and conversational conditions measured at run time, the classifier learned helps adjust the control algorithms in achieving high perceptual conversational quality. Secondly, we have designed a cross-layer speech codec to interface with the loss-concealment and playout scheduling algorithms in the packet-stream layer in order to be more robust and effective against packet losses. Thirdly, we have developed a distributed algorithm for equalizing mutual silences and an overlay network for multi-party VoIP systems. The approach leads to multi-party conversations with high listening only speech quality and balanced mutual silences.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.