Skip to main content

Table 2 Summary of calibrated models and their performance.

From: A study of artificial speech quality assessors of VoIP calls subject to limited bursty packet losses

SQA algorithm a b ρ Δ
E-Model(1) 1.170 -23.016 0.91 4.738
E-Model(2) 0.607 9.066 0.91 4.664
Genome 0.821 -2.740 0.89 5.324
VQmon 0.965 -11.694 0.87 5.741
Q-Model(1) 1.017 -11.466 0.92 4.380
Q-Model(2) 0.872 -1.344 0.92 4.473