Features
- Only 5 lines of code is sufficient to start evaluating your audio files via real humans.
- We provide hassle free human evaluation: evaluator pre-screening & training, attention test, automatic rejection of unreliable evaluations, computing statistics, generating the output report, and many more.
- You will get the evaluation report within 12 hours. Typically much faster!
Getting Started
Let’s get started.Quickstart
Start your speech & audio evaluation by human in 5 minutes.
Evaluation Analysis
Example of the final evaluation analysis
Naturalness measurement
Evaluate the naturalness of synthesized speeches.
Voice similarity
Evaluate the similarity of two voices.
Preferences
Measure preferences among the two models
Quality evaluation
Evaluate the speech quality following ITU-T P.808
SDK Reference
Reference for SDK