FAQ from SpeechEvalPro
What is SpeechEvalPro?
SpeechEvalPro is a pronunciation evaluation and scoring API that provides high-quality, multi-dimensional assessments for Chinese and English pronunciation. It utilizes advanced voice evaluation and speech recognition technologies to deliver accurate and reliable assessments for educational purposes.
How to use SpeechEvalPro?
To use SpeechEvalPro, sign up for a free trial or select a pricing plan. Once you have access, integrate the API into your educational product or application via HTTP or WebSocket requests. The API accepts audio files in recommended formats and supports various evaluation modes, including phoneme, word, sentence, and chapter. Refer to the documentation for detailed usage instructions.
Is there an SDK available for SpeechEvalPro?
Currently, an SDK is not available. You can directly use the WebAPI, which offers streaming capabilities and is lightweight and cross-platform.
What audio formats are supported for pronunciation evaluation?
It is recommended to use audio files with a 16-bit sample size, 16K sample rate, 1 channel in opus_raw, pcm, wav, or mp3 formats. Other formats may affect the scoring results.
What evaluation modes are supported, and what are the duration and text length restrictions?
SpeechEvalPro supports phoneme, word, sentence, and chapter (paragraph) modes. The duration and text length restrictions vary by mode. For phoneme & word mode, the duration is up to 20 seconds. For sentence mode, the duration is up to 40 seconds with a text length of fewer than 300 characters. For chapter mode, the duration is up to 300 seconds with a text length of fewer than 10,000 characters. Refer to the documentation for specific details.