Speaker Diarization vs Speech-to-Text
People think speech-to-text already says who spoke. It does not. Speech-to-text writes down the words. Speaker diarization splits the sound by voice and labels each part Speaker 1, Speaker 2 and so on. A meeting script needs both.
Speaker Diarization
Working out who spoke when in a recording. Each line gets a speaker tag.
Speaker diarization works out who spoke when in a recording. It splits the audio into stretches and gives each stretch a speaker label, such as Speaker 1 and Speaker 2. It does not know people's names unless it is told. It is often paired with speech-to-text, so a transcript shows who said each line. People talking over each other is the hardest part.
Speech-to-Text (Speech Recognition)
AI that listens to speech and writes the words down.
Speech-to-text is AI that turns spoken words into written text. The model listens to the sound, cuts it into tiny slices, and works out which words were most likely said. Today's models learn from many thousands of hours of speech, so they cope with accents and some noise. Many also add commas and full stops, mark who is speaking, and give each word a time. Quality is often scored by word error rate, the share of words it gets wrong.
In one line
- Speaker Diarization
- Working out who spoke when in a recording. Each line gets a speaker tag.
- Speech-to-Text (Speech Recognition)
- AI that listens to speech and writes the words down.