Unlock instant, AI-driven research and patent intelligence for your innovation.

Voice data processing method and device

A processing method and technology of speech data, applied in speech analysis, speech recognition, instruments, etc., can solve problems such as affecting the recognition rate, losing words in recognition results, occupying energy and time, and achieving good user experience, good user experience, and recognition. quick effect

Pending Publication Date: 2019-01-04
AISPEECH CO LTD
View PDF9 Cites 20 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Regarding reliability, in the prior art, if the connection between the wake-up word and what is said later is very tight, there will be a risk of missing words in the recognition result and affecting the recognition rate
For convenience, the existing technology only relies on the communication of wake-up and identification, which cannot meet the customization of full-link dialogue, especially multi-round dialogue
If these are implemented and maintained by developers coding and maintaining this complete closed loop, it will take a lot of energy and time

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Voice data processing method and device
  • Voice data processing method and device
  • Voice data processing method and device

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0120] Example 1: Combination of Oneshot technology and DUI skill "Nickname"

[0121] User: Hello Xiaochi, let me give you a nickname Xiaohei

[0122] DUI: Okay, from now on you can say hello Xiao Hei and tell me

[0123] User: Hello Xiao Hei, what's your name?

[0124] DUI: My first name is Xiaochi, and my nickname is Xiaohei

example 2

[0125] Example 2: Combination of Oneshot Technology and Multiple Rounds of Dialogue

[0126] User: Hello Xiaochi, the weather in Suzhou today

[0127] DUI: 28°C today

[0128] User: (Hi Xiaochi,) What about tomorrow?

[0129] DUI: 30°C tomorrow

example 3

[0130] Example 3: Combination of Oneshot Technology and Mobile Assistant

[0131] Set the wake-up feedback to "ding" through the DUI platform, set ONESHOT_second time period (MIDTIME) to 500ms, and set ONESHOT_third time period (ENDTIME) to 0ms. The customization effect is similar to that of Apple Siri. At the same time, by adjusting various customization items, many different effects from Siri can be achieved.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a voice data processing method and a voice data processing device. The voice data processing method is used for a client, and comprises the steps of: receiving a first voice instruction of a user, and sending a request for starting automatic voice identification training and identification of a preset wake-up word based on the first voice instruction and the preset wake-upword; receiving a result indicating whether the awakening is successful returned by means of a server; detecting whether the user sends out a second voice instruction within a first time period according to the set first time period in response to the successful awakening; tracing a starting point of audio data of the second voice instruction to a starting point of audio data of the first voice instruction to in response to the second voice instruction which is sent by the user and detected within the first time period; and sequentially sending the first voice instruction and the second voiceinstruction detected in real time to a server side for identification from the starting point of audio data of the first voice instruction till the first time period ends.

Description

technical field [0001] The invention belongs to the technical field of voice data, in particular to a voice data processing method and device. Background technique [0002] In related technologies, the "wake-up recognition" provided by some solutions is based on its voice wake-up technology, which supports users to directly say the wake-up word and work commands together, for example: Ding dong ding dong, I want to listen to Jay Chou's song, the client will Start services such as recognition and semantic understanding directly after waking up, shortening the interaction time. The "Wake-Up Recognition Continuous Speaking" provided by other solutions is based on its voice wake-up technology, which supports the continuous expression of wake-up and recognition needs, for example: Hello Xiaodu, please help me find a coffee shop. [0003] In the process of implementing this application, the inventor found that although the above technologies can directly start recognition and sem...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G10L15/22G10L15/30G10L15/06G10L15/18
CPCG10L15/063G10L15/1822G10L15/22G10L15/30G10L2015/223
Inventor 甘津瑞张顺
Owner AISPEECH CO LTD