Speech Recognition Clean-up & Editing
|
Speech to text transcript quality continues to improve, but it still has a way to go. Good quality recordings with clear speakers and no background noise or interference can produce a 90-95% accurate transcript.
However, voice recognition/text to speech systems can't always tell the difference between their, there, and they're; to, too, and two; and Philadelphia has Segway tours, not segue tours. It can also misinterpret names, technical words, acronyms, and slang; and punctuation isn't always where it should be. There are advantages to using speech to text software as well, one of which is speed. Transcripts are completed almost in real time without having to know how to type. For clients who have created transcripts with a voice recognition (speech to text) system that need a more polished document, we proofread the transcripts against the recording and edit for spelling, grammar, punctuation, and accuracy. We make the necessary revisions and deliver a professional document. |
|
