Speech AI Preferences Test
Preference test between speech synthesis modelsLoading...Mean
Comparison Graph
Mean table
| Tag | model1_A50 | model2_f0 |
|---|---|---|
| model1 | μ: 3.13 / σ: 1.30 / CI95 : 0.14 | |
| model2 | μ: 3.13 / σ: 1.30 / CI95 : 0.14 | |
Evaluation
CreatedType
Loading...CUSTOM
Stimulus
CountRequested responsesLanguage
6410 per each audio (640 total)English (United States)
Evaluator
Total responsesValid/total evaluatorsGender ratio
64020 / 24 (83%)
MaleFemaleOther
128-
Custom evaluation query (2)
Question 1Listen to the speech/audio samples and rate which one you prefer overall.
Responses
1. A is strongly preferred2. A is preferred3. About the same4. B is preferred5. B is strongly preferred
Question 2Which audio would you prefer for American Accent?
Responses
1. A is strongly preferred2. A is preferred3. About the same4. B is preferred5. B is strongly preferred