Predicting the Quality of Synthesized and Natural Speech Impaired by Packet Loss and Coding Using PESQ and P.563 Models

Abstract

SummaryThis paper investigates the impact of independent and dependent losses and coding on speech quality predictions provided by PESQ (also known as ITU-T P. 862)a nd P. 563 models, when both naturally-produced and synthesized speech are used. Twos ynthesized speech samples generated with twod i ff erent Text-to-Speech systems and one naturally-produced sample are investigated. In addition, we assess the variability of PESQ'sand P. 563's predictions with respect to the type of speech used (naturally-produced or synthesized)a nd loss conditions as well as their accuracy, by comparing the predictions with subjective assessments. The results showthat there is no difference between the impact of packet loss on naturally-produced speech and synthesized speech. On the other hand, the impact of coding is different for the twotypes of stimuli. In addition, synthesized speech seems to be insensitive to degradations provided by most of the codecs investigated here. The reasons for those findings are particularly discussed. Finally,itisconcluded that both models are capable of predicting the quality of transmitted synthesized speech under the investigated conditions to acertain degree. As expected, PESQ achievesthe best performance overalmost all of the investigated conditions.

Evaluation of objective speech transmission quality measurements in packet-based networks

“…Five repetitions are enough to achieve satisfactorily low result dispersion and uncertainties. The results for PESQ [2,3] and 3SQM [1] are in agreement with previous experiments [6,10,11,13].…”

DeepVoCoder: A CNN Model for Compression and Coding of Narrow Band Speech

“…It is also important to measure the quality in objective terms. Perception of voice by humans is still an ongoing discussion and active research field [27]. CNN decoder output, i.e.…”

Impact of the codec and various QoS methods on the final quality of the transferred voice in an IP network

“…It has not been tested for PCM transmissions affected by packet loss. However, the recommendation itself does not prevent any user from doing this, and it was successfully used in previous experiments [6,7,8,9]. The results can also be affected by Packet Loss Concealment (PLC), implemented by Linphone.…”