🎨 Vidu has introduced Vidu S1 — a model for interactive real-time video generation with voice control.
The system allows for controlling the behavior of digital avatars through voice commands, supporting an infinite video stream. The model provides 540p resolution at 25–42 FPS on standard GPUs.
🌍 The transition to interactive streaming paves the way for creating lifelike digital avatars for video calls and virtual assistants.
👤 It is now possible to interact with an AI character like a live interlocutor, controlling its actions via voice.
Source 1: https://github.com/shengshu-ai/Vidu-S1 Source 2: https://www.vidu.com/ru/vidu-stream