US 6,556,968 B1Granted

Data terminal with speech recognition function and speech recognition system

AssigneeNEC Corporation

InventorKatsumi Shiono

Life of this patent

Estimating the term…

Loading patent…

Abstract

By an operation in an operating unit 12, voice recognition, i.e., a process of recognizing voice inputted from a microphone 17, is started. A voice recognizing unit 15 executes the process of recognizing voice inputted from the microphone 18. When the result of recognition is certified, it is displayed on a display unit 13, and a loudspeaker 18 is caused to output the recognition result in voice. To prevent erroneous operation when a wind-breaking sound generated by moving a portable telephone set a handling noise generated by taking hold of the portable telephone set afresh is inputted from the microphone 17 after the start of the voice recognition by the operation in the operating unit, recognition words for preventing erroneous operation due to noise are registered, and voice recognition is automatically started again when a recognition word for preventing erroneous operation due to noise is recognized after the recognition start.

Claims3

Dependent claims

A speech recognition system comprising:

a memory storing words and noises that are not words;
an input device for receiving words and noises that are not words;
a speech recognition unit that recognizes the words and noises that are stored in said memory from among the words and noises received by said input device, and that outputs the recognized words;
a timer that measures a first time period starting when speech recognition by said speech recognition unit is initiated;
and
a controller that, during the first time period, recommences speech recognition by said speech recognition unit when said speech recognition unit recognizes one of the noises that are stored in said memory from among the noises received by said input device during the first time period, and wherein, during the first time period, said controller continually recommences speech recognition by said speech recognition unit each time said speech recognition unit recognizes one of the noises that are stored in said memory from among the noises received by said input device during the first time period.

3 more elements

Dependencies are inferred from the claim text.