Expand description
ELARA Visual State
Video as STATE, not stream. This is NOT a video codec.
§Philosophy
Traditional video: Camera → Frames → Encode → Network → Decode → Display ELARA video: Camera → Visual State → Sync Reality → Render
Visual state captures the MEANING of what’s seen, not the pixels. When network degrades, we predict and interpolate state, not freeze frames.
§Degradation Ladder
- L0: Full visual state (keyframes + deltas + all details)
- L1: Reduced visual (lower resolution, fewer updates)
- L2: Simplified visual (face + pose only)
- L3: Symbolic visual (avatar representation)
- L4: Minimal visual (static avatar + activity)
- L5: Latent visual (last known state)
Re-exports§
pub use encoding::*;pub use face::*;pub use keyframe::*;pub use pose::*;pub use prediction::*;pub use scene::*;pub use state::*;
Modules§
- encoding
- Visual State Encoding - Wire format for visual state
- face
- Face State - Facial expression and gaze as state
- keyframe
- Keyframe and Delta Encoding - ELARA-native visual compression
- pose
- Pose State - Body pose and gestures as state
- prediction
- Visual Prediction Engine - Continuity under packet loss
- scene
- Scene State - Environment and background as state
- state
- Visual State - The core representation of visual reality