Abstract
This disclosure describes an accessibility technique, making no health claim: filtering the unintended repeated activations caused by hand tremor from discrete user input, deriving the rejection window from the physiology of the tremor rather than choosing it by feel or exposing it as a user setting.
A tremor at frequency f spaces its repeats approximately 1/f apart. The method sets the debounce window to one full tremor period, W = 1/f, and makes explicit the rule that follows: a window W rejects every repeat sooner than W, and so suppresses every tremor faster than 1/W. The uncovered edge is therefore the slow side of the band, inverting the intuition that faster tremor needs a wider window. For the classical 4-6 Hz rest-tremor band, W = 0.20 s (one 5 Hz period) covers the fast half and concedes the slow tail, preserving deliberate re-activation.
The rejection is scoped per control, not globally. A tremor burst oscillates on one target, so its repeats strike the same control: a second activation of that control within W is rejected, while activation of a different control is accepted as deliberate. A global filter cannot tell that burst from a fast deliberate two-control sequence, and discards the latter.
Also disclosed: leading-edge arming, so a burst cannot extend its own window; rejection at activation onset, causing no visual stutter; default-on placement at the application layer; and variations including adaptive sizing from a per-user tremor estimate, and application to keyboard, controller, switch and pointer input.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
ziemski, gerard, "Physiologically derived, per-control input debouncing for tremor-affected touch and key input", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/11329