Execution Evidence / nmVR Edge Control
Neuromorphic Camera And VR Stabilization
Live Jetson Orin Nano stream comparing a deterministic baseline controller with SLNN+NdFractal+HRR.
Baseline telemetry is used only for measurement, not as a neural fallback.
SLNN+NdFractal+HRR
Baseline Control
Connecting...
Tracking Fixture
Baseline Error
- px
Neural Error
- px
Live Improvement
-
Baseline Overshoot / Jitter
- px / -
Neural Overshoot / Jitter
- px / -
Validated Mean
1.86x lower
Baseline
Neural
Target
Baseline: damped spring/lead camera controller. Neural: adaptive reservoir controller with NdFractal memory and HRR correction.
Current measured result: mean tracking error is about 46% lower than baseline on this fixture.
VR Fixture
Baseline Residual
-
Neural Residual
-
Live Improvement
-
Validated p95
0.0047
FPS
-
Validated Mean
1.79x lower
Baseline
Neural
Raw Motion
Baseline: deterministic horizon damping. Neural: independent SLNN+NdFractal+HRR stabilization, with no baseline output fallback or oracle critic.
Current measured result: mean VR residual is about 44% lower than baseline; p95 residual is about 0.0047.