Text to speech
음성 합성
Also known as: TTS · voice cloning
Turning text into speech, including systems that imitate a particular voice from a short recording.
Synthetic speech began with announcement style voices and now imitates intonation and pauses. For reading picture books, pace and pausing matter most.
Systems that build a voice from a short recording are also used, as with an audiobook read in a parent's voice.
Consent and retention policy matter as much as the technology: whose voice it is, where it is used, and when it is deleted should be stated plainly.
- Quality criteriaClarity, natural pauses, and consistency when the same line is read again.
- EthicsDo not clone a voice without that person's consent.