Glossary → Transcription & speech recognition
Transcription & speech recognition
Transcript editing
Transcript editing is the human pass after automatic transcription — fixing misheard words, correcting names and terms, adjusting speaker labels and tightening punctuation.
Efficient editing is audio-linked: click any word, hear that moment, fix in place. On clean audio the pass is minutes per hour of recording; the highest-value targets are proper nouns, numbers and low-confidence highlights.
Related terms
Confidence scoreA confidence score is the speech recognizer's own estimate of how likely each word or segment is correct, typi…
Speaker labelsSpeaker labels are the per-turn tags in a transcript — 'Speaker 1', 'Interviewer', or a real name — produced b…
Punctuation restorationPunctuation restoration is the step that turns raw recognized words into readable text by inserting periods, c…
Put the term to work
Transcribe audio or video in 99 languages — speaker labels, word timestamps, captions.
Transcribe a file →