Keyword Recognition Model Selection for Voice Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing voice recognition systems fail to accurately identify user intentions and allow operations to continue despite incorrect responses, lacking the ability to stop operations based on user voice commands.

Innovation Solution

An electronic apparatus with a microphone, memory, and processor that selectively executes keyword recognition models based on operating state information, identifying and responding to specific keywords within user voice inputs to perform corresponding operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a voice recognition system provides responses based on user input, then the system can interact with users, but the system cannot stop operations when the user requests to stop through voice commands

Engineering Contradiction:
Improvevoice command controlVSAvoidoperation control accuracy
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The voice recognition system is divided into multiple keyword recognition models that operate independently. Each model is responsible for recognizing specific keywords related to different operations, including stop commands. This segmentation allows the system to accurately identify and execute stop requests without interfering with other operations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system pre-trains multiple keyword recognition models with specific keywords before actual use. These models are prepared in advance to recognize various commands including stop commands. When a user speaks, the system can immediately match the input against the pre-trained models and execute the appropriate action without delay.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If multiple keyword recognition models are executed to improve recognition accuracy, then keyword identification becomes more accurate, but system complexity increases

Engineering Contradiction:
Improvekeyword recognition accuracyVSAvoidsystem structure
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

Instead of using one complex recognition system, the patent segments the recognition task into multiple specialized keyword recognition models. Each model focuses on recognizing specific keywords, making them simpler and more accurate in their respective domains. The processor selectively executes only the relevant models based on the input, managing complexity efficiently.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The processor is designed with multi-functionality to selectively execute different keyword recognition models based on the input voice data. This universal processor can handle various recognition tasks without requiring separate dedicated hardware for each model, thus managing system complexity while maintaining high recognition accuracy.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Productivity

If keyword recognition models are selectively executed based on operating state information, then recognition efficiency improves, but the system requires more complex state management

Engineering Contradiction:
Improverecognition efficiencyVSAvoidstate management
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system maintains operating state information in advance, which records the current status of various functions and operations. This pre-maintained state information allows the processor to quickly determine which keyword recognition models are relevant without complex real-time analysis, improving recognition efficiency while keeping state management straightforward.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses operating state information as feedback to selectively execute appropriate keyword recognition models. The state information continuously reflects the current system status, and this feedback mechanism guides the processor to choose the most suitable models for the current context, optimizing efficiency without requiring complex decision-making logic.

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS11580964B2Electronic apparatus and control method thereof
Publication Date: 2023.02.14 SAMSUNG ELECTRONICS CO LTD
  • US11580964B2 patent drawing
  • US11580964B2 patent drawing
  • US11580964B2 patent drawing

AI summary

An electronic apparatus is provided. The electronic apparatus includes a microphone, a memory configured to store a plurality of keyword recognition models, and a processor, which is coupled with the microphone and the memory, configured to control the electronic apparatus, wherein the processor is configured to selectively execute at least one keyword recognition model among the plurality of keyword recognition models based on operating state information of the electronic apparatus, based on a first user voice being input through the microphone, identify whether at least one keyword corresponding to the executed keyword recognition model is included in the first user voice by using the executed keyword recognition model, and based on at least one keyword identified as being included in the first user voice, perform an operation of the electronic apparatus corresponding to the at least one keyword.