Today we're releasing our biggest accuracy improvement yet for phone call transcription, making AssemblyAI the most accurate Speech to Text API for transcribing phone calls.
In our latest release, we've trained our Speech-to-Text models to be robust to low bitrate and compressed audio data, commonly found in phone call recordings. Read more, and see how we compare to Google on phone call audio, in this update!
Custom models can help you significantly improve accuracy by learning the speaking style and linguistic patterns of the speech in your application. Follow this quick guide to learn how to build a custom model for your application.
Doing keyword spotting, or looking to detect certain phrases in a recording? Follow this quick guide to learn how you can boost accuracy for specific keywords and/or phrases using a custom model.
You can now pip install assemblyai! The module is a thin Python 2/3 wrapper to make it easier to integrate the API. Learn more in this post.