Inventor(s)

Abstract

A dictation app gives a single global key a small set of gestures, each with distinct feedback in its recording HUD. Holding the key records; on release the speech is transcribed and inserted into the application that was in front when the key went down. Audio under 300 milliseconds, silence, a press before the models are ready, or a change of front application each produce a specific message instead of text. Two short taps take back the last dictation with guarded Backspaces. Pressing the translation key or the voice-command chord during a recording relabels that same recording, without restarting it, as a translation or a command, and releasing that key ends it. Pressing the mouse on the HUD capsule during a recording abandons the recording and turns the capsule into an up-or-down selector for two translation features. The gestures are designed not to collide: a hold is never a tap, a tap records too little to produce text, a relabelled recording keeps its audio, and a pointer takeover discards the audio before anything is inserted.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS