Visemator: AI-Powered High-Precision Lip Sync Plugin (Supports ARKit 52BS)

Bring lifelike speech synchronization to your virtual characters.

Visemator is a powerful Unreal Engine plugin designed for virtual humans, animation production, and real-time interactive scenarios. Powered by deep learning, it accurately converts Mandarin and English audio into 52 ARKit-standard BlendShape (BS) weights, enabling natural, fluid viseme animation.

🔹 Key Features

• Bilingual Audio Support: Supports both Chinese and English speech, covering the most common use cases.

• AI-Trained Precision: Built on hundreds of hours of audio-viseme paired data, delivering smooth, expressive, and natural lip sync results.

• Ultra-Fast Processing: Achieves millisecond-level inference speed, ideal for both real-time applications and offline rendering.

• Plug-and-Play: Fully integrated with Unreal Engine, including a ready-to-use demo project for quick testing and deployment.

• Seamless Compatibility: Easily connects with animation blueprints, MetaHuman framework, and virtual character pipelines.

• LiveLink Requirement: Requires the Unreal Engine LiveLink plugin to be enabled for real-time animation streaming.

🎯 Ideal For

• Digital humans, virtual streamers, AI voice assistants

• Game character lip-sync and dialogue performance

• Digital twin systems, XR interaction, virtual meetings

• Animation films, short-form content, and educational media

📦 Demo Download

• Google Drive – Visemator Demo

🎥 Video Demonstrations

• Visemator UE Plugin – Fast Audio-Driven Lip Sync with MetaHuman

• Visemator UE Plugin – Fast Lip Sync with Custom 3D Characters

📖 Online Documentation

• Visemator User Manual (PDF)

🌐 Official Website

• https://www.visemator.com