Keyboard and button preferences for processing multimodal inputs

A multimodal device, such as a PDA or cellphone supports multiple inputs for communication between the user and the application. Multiple inputs include keyboard, touch or tap screen, hand writing recognition, and voice recognition. The entered data can be both displayed and spoken back to the user.

Page 1 of 2

Keyboard and button preferences for processing multimodal inputs

  A small device will have several hardware buttons that can be used to activate speech recognition for data input. A touchpad keyboard key, accessed with a stylus, may also be used. For a PC, the user would select a keyboard key.

There are four different listening modes for processing speech recognition input The user would select one of these modes to determine how to begin and end speech recognition with the selected access button or key:

Push-to-talk, depress button to begin talking, release button to end talking. Push-to-talk, depress and release button to begin talking and depress and release button to end talking. Push-to-activate, depress and release button to begin talking, recognition detects silence to end talking. Always on, the recognition is always listening, button is not depressed.

The user sets preferences that determine which button to press, and optionally release, and the listening mode to use when speaking into the device. The user may have a specific button preference based on being right- or left-handled, for example. The user preferences may either be set globally for all applications installed on the device, or for each application individually. The individual application preferences will override the global device preference settings.

The optimal listening mode would apparently be always on, where a speech recognition activation button is not needed. However, this listening m...