Dynamic Speech End Detection Period Adjustment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The accuracy of speech recognition in electronic devices is compromised when the utterance end detection period is fixed, leading to misrecognition of sentences due to either incomplete recognition with short detection periods or delayed output with long detection periods.

Innovation Solution

An electronic device with a microphone, processor, and memory that dynamically adjusts the speech end detection period based on identified pause periods between speech stop and resume time points, determining the end of speech input accordingly.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If the utterance end detection period is set to be short, then the output speed of recognition result is improved, but the speech recognition accuracy deteriorates due to failure to recognize complete sentences

Engineering Contradiction:
Improveoutput speed of recognition resultVSAvoidspeech recognition accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The patent applies dynamics by making the utterance end detection period adjustable rather than fixed. The system dynamically changes the detection period length based on pause period information, allowing it to adapt between short periods (for faster output) and long periods (for higher accuracy) depending on the specific speech context and user needs.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameter of detection period length based on pause period analysis. By analyzing pause periods in speech and using this information to adjust the detection period parameter, the system optimizes both recognition accuracy and output speed for different speech patterns and contexts.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If the utterance end detection period is set to be long, then the speech recognition accuracy is improved, but the output speed deteriorates due to delayed recognition result output

Engineering Contradiction:
Improvespeech recognition accuracyVSAvoidoutput speed of recognition result
Core Design Contradiction:
Measurement precisionVSSpeed

Solution Approach 1:

The system dynamically adjusts the detection period length based on analyzed pause period information rather than using a consistently long period. This allows the system to achieve high accuracy when needed while maintaining faster output when the speech context allows.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The detection period parameter is changed based on pause period analysis results. The system uses pause period information to determine appropriate detection period lengths, optimizing the balance between accuracy and speed for different speech scenarios.

Inventive Principle:
Principle #35Parameter changes

3Device complexity

If a fixed detection period is used, then the device complexity is reduced, but the adaptability to different user speech habits deteriorates

Engineering Contradiction:
Improvedetection system complexityVSAvoidadaptability to user speech habits
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The system performs self-service by automatically analyzing pause periods in speech and adjusting the detection period length without requiring manual user configuration. This enables the system to adapt to different user speech habits autonomously while maintaining relatively simple device architecture.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system uses feedback from pause period analysis to adjust the detection period parameter. By continuously analyzing speech pause patterns and using this feedback to optimize detection timing, the system adapts to user speech habits while maintaining system simplicity.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3391367B1Electronic device and speech recognition method thereof
Publication Date: 2020.12.02 SAMSUNG ELECTRONICS CO LTD
  • EP3391367B1 patent drawingFigure 1
  • EP3391367B1 patent drawingFigure 2
  • EP3391367B1 patent drawingFigure 3

AI summary

An electronic device and a speech recognition method that is capable of adjusting an end-of-utterance detection period dynamically are disclosed. The electronic device includes a microphone, a display, an input device formed as a part of the display or connected to the electronic device as a separate device, a processor electrically connected to the microphone, the display, and the input device, and a memory electrically connected to the processor. The memory stores instructions, executable by the processor, for receiving an utterance input by a user through the microphone, converting the utterance to text comprised of a series of words or phrases with spaces, displaying the text on the display, the text comprising at least one space formed at an incorrect position, and receiving a user input for updating a predetermined time period through the input device.