Spontaneous physiological blinks and voluntary blinks must be distinguished

What it is

A blink is not one kind of event. A spontaneous blink wets the cornea on its own, a dozen to twenty times a minute while awake, and people often do not remember it. A voluntary blink is an intentional lid closure used as a command. Treating every closure as input conscripts a corneal reflex as a key, and false positives will fill ordinary looking. To be input, the two have to be split first. This is not dwell (which integrates fixation duration) and not Midas (which conscripts looking).

Why it happens

Spontaneous blinks are driven by the brainstem and the tear film, often 100–150 ms, both lids together, irregularly spaced, more frequent when tired or dry. Voluntary blinks are cortically launched; people can make them longer, harder, sometimes on one side only. A classifier can use duration, laterality, and whether the event falls in a prompted window—not “was there a closure,” which carries almost no information.

The distributions overlap. Stress or wind lengthens spontaneous blinks; a voluntary blink performed as an ordinary wink falls into the spontaneous duration band. “Blink once to click” is therefore almost inseparable in mechanism; what can be separated is voluntarily stretching the closure, adding laterality, or adding an explicit preparatory pose.

Studying it

Log in parallel the tracker’s lid signal, palpebral fissure from video, and a key the participant presses to say “that one was on purpose.” Compare classifier precision and recall across a spontaneous baseline, a dryness exposure, and a condition that asks for a blink whenever a target appears. Outcomes must include false positives (spontaneous counted as command) and misses (voluntary not recognized). Instructions change spontaneous rate: once people know blinks are input they suppress them, and laboratory separability beats the product. Dry-eye and glasses wearers need their own measurements; their spontaneous durations are longer.

Where it stops holding

Parkinson’s, facial palsy, and hemifacial spasm scramble bilaterality and duration; the boundary must be redrawn or the blink channel abandoned. Infants have almost no usable “voluntary blink.” In some cultures a deliberate blink is social, and as a command it will be misread in public. Communication boards that use blink as the only input must accept a higher false-positive rate plus undo, not assume perfect classification.

Applying it

• Do not bind every detected closure to a command; classify on duration and bilaterality first, and only events that leave the spontaneous band enter command interpretation.
• Provide an on/off for blink input in settings, default off for ordinary browsing.
• Verify by recording quiet viewing with no task and counting how many spontaneous blinks the current rule would treat as commands; then record prompted voluntary blinks and count misses.

Related

• Same group: C8.11.2 Voluntary blink commands are usually recognized by lid-closure duration, not a single blink · C8.11.3 Using a wink as a command requires left/right discrimination and is infeasible for some users · C8.11.4 False triggers of blink input are easily confused with fatigue and dryness
• Adjacent: C8.02 Dwell selection · C8.03 Midas touch in eye input
• Search: spontaneous blink · voluntary blink · blink as input