How it Works

Capture

• Performers are scanned in-studio using our LightCage system or mobile via a self-scan app or a remote scan rig.

• Scan captures look, performance training, and movement data

• This data trains:

  •  a 3D model based on SIMPL that not only understands the actors look, but how they move

  • A face identity model that learns their identity and emotional performance and is adaptable to emotion in a scene

  • A series of custom LoRas that understand their general look 

• Creates a consented digital identity with embedded rights and ownership

• Stored in AI-safe performance vault

Performance

• The performer acts at home using an iPhone or in a studio using a multi-camera setup for larger action

• System tracks facial performance and full body movement

• Additional action (stunts, runs, jumps) can be prompted and added, and blended on a timeline

Deployment

• 3D performance placed in any scene with any camera angles, or exported to another 3D program for editing

• Rendered with AI using reference for styling (makeup, hair, wardrobe, environment)

• AI rendering uses 3D geometry-guided rendering to convert the nuance of the 3D mesh to the diffuser rendered asset

• Can render full environment or ‘performance as an asset’ for traditional compositing