This dataset contains all 11,000+ ratings collected for 20 synthetic speech systems for the TTSDS2 study (link coming soon).
The scores are MOS (Mean Opinion Score), CMOS (Comparative Mean Opinion Score) and SMOS (Speaker Similarity Mean Opinion Score).
All annotators included passed three attention checks throughout the survey.
