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

Awakening method and device for intelligent equipment based on artificial intelligence

A smart device and wake-up word technology, applied in the field of data processing, can solve problems such as false wake-up of smart devices, high false wake-up rate, poor user experience, etc.

Pending Publication Date: 2020-03-17
TENCENT TECH (SHENZHEN) CO LTD
View PDF0 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, the current related technologies have a high false wake-up rate, that is, some noise or non-wake-up word voices are misidentified as wake-up word voices, and wrongly wake up the smart device, causing the smart device to suddenly start when the user does not need it. , resulting in a bad experience for users

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
  • Awakening method and device for intelligent equipment based on artificial intelligence
  • Awakening method and device for intelligent equipment based on artificial intelligence
  • Awakening method and device for intelligent equipment based on artificial intelligence

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0051] Embodiments of the present application are described below in conjunction with the accompanying drawings.

[0052]In the existing smart device wake-up technology, single-model custom wake-up technology is often used, that is, only one audio recognition method is used for audio recognition, for example, only Keyword / Filler Hidden Markov Model (Keyword / Filler Hidden Markov Model, referred to as HMM ) or only use the Long Short-term memory Feature Extractor System (LSTM Feature Extractor System for short) for audio recognition, and no matter what kind of audio recognition technology, when the smart device wakes up alone, there are certain False wake-up rate, and the false wake-up rate is relatively high.

[0053] For example, the wake-up word of the smart device is "turn on the speaker". Since the smart device can collect audio data at any time, if the user is chatting with other people and utters the voice "turn on the music", then the smart device can collect the audio d...

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 embodiment of the invention discloses an awakening method and device for intelligent equipment, and the method comprises the steps: storing acoustic features determined according to to-be-recognized audio data in a process of verifying whether the to-be-recognized audio data contains an awakening word corresponding to intelligent equipment or not after the intelligent equipment collects the to-be-recognized audio data; after the first-level verification determines that the to-be-identified audio data contains awakening words, not waking up the intelligent equipment, but firstly performingsecond-level verification, determining an undetermined feature sequence from the stored acoustic features, and determining whether the undetermined feature sequence meets a wake-up condition or not according to the acoustic feature sequence of the awakening words; and after determining that the undetermined feature sequence meets the wake-up condition, the to-be-identified audio data passes the first-level verification and the second-level verification, and the intelligent equipment is awakened at the moment. The acoustic characteristics determined by the first-level verification are utilized,and the to-be-determined characteristic sequence is extracted according to the acoustic characteristics to perform the second-level verification, so that the error awakening frequency of the intelligent equipment is effectively reduced.

Description

technical field [0001] The present application relates to the field of data processing, in particular to a method and device for waking up smart devices based on artificial intelligence. Background technique [0002] At present, smart devices are becoming more and more popular and widely used in people's work and life. [0003] Some smart devices are in a dormant state when not providing services. When users need to use such smart devices, they can wake up the smart devices by uttering a wake-up word through voice. For example, users can wake up a dormant smart speaker through the wake-up word. [0004] However, the current related technologies have a high false wake-up rate, that is, some noise or non-wake-up word voices are misidentified as wake-up word voices, and wrongly wake up the smart device, causing the smart device to suddenly start when the user does not need it. , resulting in a bad user experience for users. Contents of the invention [0005] In order to sol...

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
IPC IPC(8): G10L15/22G10L15/26
CPCG10L15/22G10L15/26G10L2015/226
Inventor 陈杰苏丹金明杰朱振岭
Owner TENCENT TECH (SHENZHEN) CO LTD