SpeechSuper - Speech Assessment API
p/speechsuper-speech-assessment-api
Empower language Learning through Speech assessment API
Qiusi Dong
SpeechSuper β€” APIs to assess speaking skills in language learning
Featured
4
β€’
SpeechSuper provides cutting-edge AI speech assessment APIs for language learning products.

Comprehensive feedback covers pronunciation, fluency, completeness, rhythm, stress, liaison, etc.
Replies
Best
Qiusi Dong
Hi Product Hunt πŸ‘‹ We’re excited to share SpeechSuper with you! Back in 2017, Maker Xiaofei and I both worked on an English learning App featuring various speaking activities with AI-based instructive feedback. We saw so many cases where users' speaking skills make progress after regular practice. We firmly believe that speaking is essential to language learning, and feedback is essential to speaking activities. Hence comes SpeechSuper. SpeechSuper APIs assess users' audio and give instructive feedback on spoken language activities in language learning. See what SpeechSuper APIs offer: βœ… Phoneme-level pronunciation scores for spoken words & word-level pronunciation scores for spoken sentences and paragraphs, plus completeness, fluency, and speed βœ… Give feedback in real-time, making the user experience more interactive βœ… Detect mispronunciation, syllable stress for spoken words βœ… Spelling and phoneme alignment of words, useful for phonics learning βœ… Detect linking and end-of-sentence tone for spoken sentences SpeechSuper APIs support English, Mandarin Chinese, German, French, Korean, Japanese, Russian, Spanish, and more languages to come. πŸ‘‰ Try demos here (PC browser only):www.speechsuper.com/demo/english... πŸ‘‰ Learn more here: www.speechsuper.com πŸ”₯ You can use SpeechSuper APIs for FREE for 1 month! πŸ”₯ Please send an email to ph@speechsuper.com to apply one-month free trial! Preferably please include your name, company, role, and languages of interest in your email, so that we can learn more about your needs. πŸ‘‚ Xiaofei, I and the rest of the team would love to hear your feedback!
Charlie  Sheng
Cool! What's the algorithm behind the assessment?
Xiaofei Yang
@charlie_sheng We use latest deep learning methods to train acoustic models(conformer based), and then apply algorithms similar to GOP(Goodness of Pronunciation) but with much more other features to get phoneme score. Word and sentence level score are calculated based on phoneme score.
Xiaofei Yang
We are so excited to be live! We would love to hear what you think about SpeechSuper.