Neurovision · Perception & optics
A computational architecture for perception and latency, directed toward future intelligent lenses, with ethics integrated by design and documentary checks against published literature.
What it is
Neurovision models the biological visual processing chain — from retina to perception — and translates it into an intelligent-lens architecture. The model is not assumed: the lab re-derives its timings and stages against published scientific literature and corrects the system when the evidence demands it.
The ethics layer is a structural condition of the design: augmented perception demands explicit limits, not afterthought patches.
Current stage: computational model under development and documentary review; there are no sensors, materialized lenses, or physiological or clinical validation yet.
Public objective: translate a rigorous understanding of visual perception into intelligent lenses that respect privacy and autonomy.
Next verifiable milestone: complete the derivation of the perceptual chain and its intervals before fixing hardware requirements.
Lab updates
-
Documentary check
July 2026 · week 1
Chain of biological latencies re-derived against the literature
The local technical review detected a conceptual labeling error in the visual system's timing model: the ~13 ms chain corresponds to subcortical axonal conduction, while reportable conscious perception occurs around ~300 ms — consistent with published neuroscience literature. The model was corrected and its interfaces preserved the required references; this check corrects the computational representation without demonstrating original neurophysiology.
-
Provisional hypothesis
August 2026 · week 2
Coherence weights are not calibrated yet
CSI-17 remains a provisional heuristic: its formulation is traceable, but it lacks empirical calibration. It is not elevated to a constant or to neurophysiological evidence, and it does not by itself enable sensor, lens, or hardware requirements.
Literature checks and interface tests are reported separately from any physiological experiment, sensor, or hardware. Human adjudication bounds scope. The internal mechanism is never published.