The study cross-validates the magnitude estimation (ME) procedure and the pair comparisons (PC) procedure for use in scaling the quality of text-to-speech synthesis. To this goal the psychophysical scale values obtained from the ME procedure were compared to the values obtained from the PC procedure. In the PC procedure the subject indicates his preference for one of two stimuli. The ME procedure requires the subject to make direct numerical estimations of the sensory magnitudes produced by different stimuli. Four different synthesis systems and three different prosodic rules were scaled. The results indicate a good general agreement between the results of the two procedures.
Cite as: Pavlovic, C.V., Sorin, C., Roumiquiere, J.P., Lucas, J.P. (1989) A comparative analysis of the magnitude estimation and the pair comparison techniques for use in assessing quality of text-to-speech synthesis. Proc. Speech Input/Output Assessment and Speech Databases, Vol.2, 91-93
@inproceedings{pavlovic89_sioa,
author={Chaslav V. Pavlovic and Christel Sorin and Jean Pierre Roumiquiere and Jean Pierre Lucas},
title={{A comparative analysis of the magnitude estimation and the pair comparison techniques for use in assessing quality of text-to-speech synthesis}},
year=1989,
booktitle={Proc. Speech Input/Output Assessment and Speech Databases},
pages={Vol.2, 91-93}
}