Neuroadaptive Interface 2D: EEG Band-Power Attention Classifier

Mental State

Custom beta / alpha bias 0.50

Signal noise level 0.15

Classifier Thresholds

Focus threshold (engagement) 1.00

Fatigue threshold 1.00

RELAXED

Engagement (β/(α+θ))

0.00

Fatigue ((θ+δ)/(α+β))

0.00

Dominant band

—

Band Power (filtered)

How it works

This is the 2D counterpart to the 3D neuroadaptive-interface scene: five synthetic cortical oscillators (delta, theta, alpha, beta, gamma) are summed into one raw scalp signal, then decoded back out with five real 2nd-order IIR bandpass (RBJ cookbook) filters at 256 Hz — the same DSP a real EEG front-end runs before band-power estimation:

H(z) = (b0 + b1 z⁻¹ + b2 z⁻²) / (1 + a1 z⁻¹ + a2 z⁻²)
power_band(t) = EMA[ y_band(t)² ]

Two ratiometric indices are computed from the five band powers, exactly as in the 3D scene:

engagement = P_beta / (P_alpha + P_theta + ε)
fatigue    = (P_theta + P_delta) / (P_alpha + P_beta + ε)

This 2D version replaces the mock adaptive UI panel with an explicit threshold-based state classifier: engagement is compared against a Focus threshold and fatigue against a Fatigue threshold to output one of three discrete states — FOCUSED, FATIGUED or RELAXED — the same decision a real attention-classifying BCI makes before triggering an interface change.

Mental state / mix — sets the target amplitude of each oscillator band the raw signal is built from.
Noise level — additive white noise mixed into the raw signal before filtering, degrading the band estimates exactly as electrode/movement noise does on real EEG.
Eye-blink artifacts — periodic large low-frequency transients, the classic EEG contamination source, injected into the raw signal.
Focus / Fatigue thresholds — the decision boundaries the classifier compares the live engagement and fatigue indices against; raise a threshold to require a stronger signal before that state fires.

Raw scalp signal Beta / engagement