Electronic Device Wakeup Using MFCC Message Intermediary

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In networked electronic devices, such as smart home devices, the slow wakeup responses due to poor arithmetic capabilities and network delays lead to inefficient overall system performance and user experience, as devices with low processing capabilities take longer to wake up, affecting the entire group's efficiency.

Innovation Solution

An electronic device with high calculation capability is woken up first to send a broadcast wakeup message containing MFCC-based instructions, allowing devices with low processing capabilities to wake up directly based on the message, bypassing the complex voice instruction processing, thus improving overall wakeup efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If devices with low processing capabilities execute the complete voice instruction wakeup mechanism, then they can independently wake up, but the wakeup response time becomes unacceptably long

Engineering Contradiction:
Improveindependent wakeup capabilityVSAvoidwakeup response time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent introduces a high-performance device as an intermediary that processes voice instructions and generates wakeup messages. Low-performance devices receive these pre-processed messages instead of handling raw voice instructions, significantly reducing their processing time while maintaining independent wakeup capability through the simplified message verification process

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The high-performance device performs preliminary processing of voice instructions by extracting MFCC features and generating wakeup messages in advance. This preliminary action transfers the computational burden to a capable device, allowing low-performance devices to quickly verify and respond to pre-processed wakeup messages without executing the complete voice processing pipeline

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If all devices execute the complete voice instruction processing pipeline, then accurate wakeup decisions can be made, but the overall system wakeup efficiency deteriorates

Engineering Contradiction:
Improvewakeup decision accuracyVSAvoidoverall system wakeup efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the wakeup processing pipeline into two distinct stages: (1) voice instruction processing and MFCC extraction performed by high-performance devices, and (2) simplified wakeup message verification performed by all devices. This segmentation allows each device to execute only the necessary portion of the processing pipeline, maintaining accuracy while improving overall system efficiency

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Low-performance devices perform only partial processing by verifying wakeup messages against their local voiceprint databases, rather than executing the complete voice instruction processing pipeline. This partial action is sufficient for wakeup decisions since the heavy lifting of voice feature extraction has already been done by high-performance devices, thereby improving system-wide wakeup efficiency

Inventive Principle:
Principle #16Partial or excessive action

3Adaptability or versatility

If devices with poor arithmetic capabilities process voice instructions locally, then they can make wakeup decisions, but the decision speed becomes unacceptably slow

Engineering Contradiction:
Improvelocal wakeup decision capabilityVSAvoidwakeup decision speed
Core Design Contradiction:
Adaptability or versatilityVSSpeed

Solution Approach 1:

High-performance devices act as intermediaries that preprocess voice instructions and generate wakeup messages containing extracted MFCC features. Low-performance devices receive these pre-processed messages and perform only lightweight verification against their local voiceprint databases, maintaining local decision capability while dramatically improving decision speed

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates simplified copies of voice instructions in the form of wakeup messages that contain essential MFCC features and device identification information. Low-performance devices verify these copied representations against their local voiceprint databases rather than processing original voice instructions, preserving decision capability while accelerating the process

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3923275B1Device wakeup method and apparatus, electronic device, and storage medium
Publication Date: 2023.12.27 BEIJING XIAOMI PINECONE ELECTRONICS CO LTD
  • EP3923275B1 patent drawingFigure 1~2
  • EP3923275B1 patent drawingFigure 3~4
  • EP3923275B1 patent drawingFigure 5~6

AI summary

The present disclosure relates to a device wakeup method and apparatus, an electronic device, and a storage medium. The wakeup method is applied to a first electronic device and includes: a wakeup message from a second electronic device is received, and when determining that a present state is an unawakened state, locally collected voice data is acquired (S11); MFCC extraction is performed on the voice data to acquire a first MFCC of the voice data (S12); the wakeup message is parsed to obtain a second MFCC comprised in the wakeup message (S13); the first MFCC and the second MFCC are matched, and when determining that a difference between the first MFCC and the second MFCC is less than or equal to a set threshold value, a wakeup instruction is generated (S14); and responsive to the wakeup instruction, the first electronic device is woken up (S15).