YouTube is working on a new AI lip-sync feature that makes automatically dubbed videos look more natural by matching mouth movements to translated audio. The system uses pixel-level facial modeling to adjust lips, teeth, and expressions for realistic results, according to YouTube’s Auto-dubbing lead Buddhika Kottahachchi.
Initially, it will support 1080p videos and a few languages — English, French, German, Portuguese, and Spanish — with plans to expand to over 20. The rollout will begin with a small pilot group, similar to YouTube’s earlier auto-dubbing launch, and may include an optional paid tier.
YouTube will add AI-use disclosures and invisible watermarks to prevent misuse. Other platforms, including Meta’s Instagram, are also testing similar multilingual lip-sync tools.
Source: Android Authority
