Skip to content
Live
Trending storyPast story

Wizard model achieves real-time world manipulation

1 reports1 outlets17 days agoUpdate

The story so far

AI overview

Hugging Face team releases Wizard model, which decouples character animation, scene control, and game logic from visual generation, enabling the intelligent agent to control joint skeletons and scene geometry in real-time, achieving 24 FPS interaction, and supporting zero-shot generalization to unseen scenes, characters, and actions. The team built a 50.3-hour agent-video pair dataset containing 750 scenes and 178 characters, using character adaptive masking, causal adaptation, and few-step distillation techniques to achieve real-time interaction and generalization capabilities. The Wizard model, through structured visual proxy control, allows the intelligent agent to manipulate any world in real-time without retraining for new scenes, characters, or actions. Technical details and code for the model are about to be open.

AI-written from the reports · Updated 17 days ago

Coverage timeline

Follow the coverage to see every side of the story.

2 Nov
  1. Hugging Face Daily Papers
    Wizard: Manipulate Any World in Real Time via Agentic Control

    Wizard decouples character animation, scene control, and game logic from visual generation, allowing intelligent agents to control joint skeletons and scene geometry through structured visual agents, achieving real-time interaction at 24 FPS, zero-shot generalization to unseen scenes, characters, and actions. The team built 50.3 hours of aligned agent-video pairs, including 750 scenes and 178 characters, supporting character adaptive masking, causal adaptation, and few-step distillation techniques.