In one scenario, Microsoft suggests the system could be used to customize a character's voice in a film or game, or to replace a user's voice with a synthetic version across multiple formats. The voices of actors, podcasters, or politicians could be preserved and translated across multiple languages using AI. The system could also be used to convert text to speech in apps and bring virtual personalities to life. However, Microsoft's filing does not state the intended purpose of the invention.
Key takeaways:
- Microsoft has filed a patent for an 'automatic dubbing' system that can remove speech from media and replace it with new voices.
- The system can clone people's voices, translate them into different languages, and can be used on personal devices, websites, DVD players, and more.
- Microsoft suggests that this technology could reduce dubbing costs for entertainment and media companies by translating a voice actor's speech into different languages, instead of hiring human linguists.
- The technology could also be used to preserve and translate the voices of actors, podcasters, or politicians across multiple languages to reach a wider audience using AI.