Voice identification system

Abstract

The invention discloses a speech recognition system, which comprises at least one speech recognition engine and a display device. The display device is provided with a signal indication interface and a character output interface, wherein the signal indication interface displays a recording state, a speech recognition ongoing state or a state of finishing the speech recognition by means of a waveform; and the character output interface is used for displaying a character result of the speech recognition, wherein each waveform unit on the signal indication interface and each work unit on the character output interface are connected with a feedback adjustment option group respectively to select an instruction for a user to modify a wrong recognition result and adjust the speech recognition system. The system is applicable to speech recognition systems of electronic devices with graphic display interfaces such as desktop computers, notebook computers, home multimedia systems, televisions, DVDs, audio and video systems, mobile phones and personal digital assistants.

Description

technical field [0001] The invention relates to the field of speech recognition, in particular to a method for users to select instructions to correct speech recognition errors or adjust a speech recognition system. Background technique [0002] In terms of the use of many electronic devices today, the development of speech recognition technology provides users with a more convenient path. Regardless of desktop computers, notebook computers, mobile phones, personal digital assistants or other electronic devices, their input must rely on the cooperation of vision and body to achieve the purpose of manipulation. For example, when a user controls a computer, he must rely on the assistance of a keyboard, mouse or other auxiliary control devices to complete the input of instructions. Or use a touch screen to simplify the input process. However, due to the limited display area of ​​the screen and the need to use fingers to operate, the maximum convenience cannot be achieved. The...

AI Technical Summary

Problems solved by technology

However, while displaying the status, it cannot reflect the quality of the recording or recognition results or whether the recording process is successful

[0005] In addition, for the results of speech recognition, the conventional technology may provide some adjustment functions that can be made according to the recognition results, but most of them only need to be set for the overall speech recognition results, so it is often impossible to only target a certain part of the speech recognition results Make adjustments to feed back and sharpen the functionality of the speech recognition system to better suit individual user needs

For example, individual users have special accents in the pronunciation of certain words or words. If the speech recognition results of these words and words cannot be fed back and adjusted, it will be impossible to provide a speech recognition system suitable for individual users. thereby greatly reducing its effectiveness in practical use

Images