Keyboard apparatus and method with voice recognition
First Claim
Patent Images
1. In a text entry device having a voice input and voice recognition means, a method comprising the steps of:
- receiving a spoken word from an operator by means of the voice input;
failing by the voice recognition means to recognize the spoken word to within a predetermined degree of accuracy;
selecting a most likely candidate word for the spoken word;
alerting the operator of the most likely candidate word;
receiving from the operator one of;
a confirmation indicating acceptance of the most likely candidate word; and
an alternate word indicating rejection of the most likely candidate word;
if the confirmation is received from the operator, converting the most likely candidate word into keystroke data representing the constituent characters of the most likely candidate word;
if the alternate word is received from the operator, converting the alternate word into keystroke data representing the constituent characters of the alternate word; and
outputting the keystroke data.
1 Assignment
0 Petitions
Accused Products
Abstract
A standalone keyboard (101) for use with computers (102), which keyboard (101) has integral voice recognition capabilities to facilitate both text entry and cursor-movement control. The speech recognition platform (204) may be speaker dependent and/or trainable. The keyboard (101) may have an optional display (112) to facilitate certain training activities. Recognized words are converted (206) in the keyboard (101) into keystroke data that corresponds to keystroke data as would have been output had the corresponding text been entered by an operator using the keyboard keys (107).
73 Citations
4 Claims
-
1. In a text entry device having a voice input and voice recognition means, a method comprising the steps of:
-
receiving a spoken word from an operator by means of the voice input; failing by the voice recognition means to recognize the spoken word to within a predetermined degree of accuracy; selecting a most likely candidate word for the spoken word; alerting the operator of the most likely candidate word;
receiving from the operator one of;a confirmation indicating acceptance of the most likely candidate word; and an alternate word indicating rejection of the most likely candidate word; if the confirmation is received from the operator, converting the most likely candidate word into keystroke data representing the constituent characters of the most likely candidate word; if the alternate word is received from the operator, converting the alternate word into keystroke data representing the constituent characters of the alternate word; and outputting the keystroke data. - View Dependent Claims (2, 3, 4)
-
Specification