Glossary term
Glossary term
Multimodal AI
AI technique that creates a synthetic voice model from a small number of reference audio samples.
ElevenLabs' voice cloning creates a personalised voice model from 1 minute of audio - used by the Stephen Hawking Foundation to recreate Hawking's synthesised voice for new content from his archived recordings.
Respeecher is used in film post-production to replace damaged audio or generate dialogue for de-aged actors - used in The Mandalorian to recreate young Luke Skywalker's voice from archival recordings.
HeyGen's avatar product uses voice cloning to lip-sync translated video content - a CEO records a video in English and HeyGen generates versions in Spanish, French, and Mandarin with the CEO's cloned voice.