Voice Input Device Dynamic Period Adjustment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing voice input systems face challenges in accurately determining the voice activity period, leading to inaccuracies in voice recognition, especially when the user's operation on the touch icon does not synchronize with the utterance timing, resulting in premature or delayed end points of voice activity periods.

Innovation Solution

A voice input device and method that include an input unit, storage unit, and controller, which acquire and store voice data and adjust the voice activity period based on user operations on an operation unit, allowing real-time adjustment of the end point through slide operations to synchronize with the utterance timing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If the voice activity period is determined based on a simple switch operation, then the ease of operation is improved, but the measurement precision of the voice activity period deteriorates

Engineering Contradiction:
Improveease of operationVSAvoidmeasurement precision
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent makes the voice activity period dynamically adjustable after initial determination. The controller allows users to modify the determined period through subsequent operations, transforming a static determination into a dynamic adjustment process that resolves the contradiction between simple operation and precise measurement

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system provides feedback by displaying the determined voice activity period to the user and allowing modifications based on actual utterance timing observation. This feedback loop enables users to refine the measurement precision while maintaining ease of operation through intuitive adjustme

Inventive Principle:
Principle #23Feedback

2Productivity

If the voice activity period is determined immediately upon switch operation, then the productivity is improved, but the reliability of voice recognition deteriorates

Engineering Contradiction:
ImproveproductivityVSAvoidreliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary determination of the voice activity period based on the switch operation, then allows for subsequent adjustment. This preliminary action maintains productivity by providing immediate functionality while the adjustment capability ensures reliability by correcting any timing inaccuracies before final recognition

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The voice activity period transitions from a fixed predetermined value to a dynamically adjustable parameter. The controller enables modification of the period based on actual utterance timing, allowing the system to adapt between speed of operation and accuracy of recognition timing

Inventive Principle:
Principle #15Dynamics

3Reliability

If the detection range is set wider to capture the beginning of utterance, then the reliability is improved, but the measurement precision of the end point deteriorates

Engineering Contradiction:
ImprovereliabilityVSAvoidmeasurement precision
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent segments the voice activity period determination into two distinct phases: initial detection phase that uses a wide range to reliably capture the utterance beginning, and refinement phase that precisely determines the end point. This segmentation allows each phase to optimize for its specific purpose without compromise

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The detection range dynamically changes during the determination process. The controller initially uses a wider detection range to ensure reliable capture of the utterance start, then transitions to a more precise measurement mode for determining the end point, adapting the range based on the determination stage

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS10854200B2Voice input device, translation device, voice input method, and recording medium
Publication Date: 2020.12.01 PANASONIC INTELLECTUAL PROPERTY MANAGEMENT CO LTD
  • US10854200B2 patent drawing
  • US10854200B2 patent drawing
  • US10854200B2 patent drawing

AI summary

A voice input device includes an input unit, a storage unit, an operation unit, and a controller. The input unit acquires a voice and generates voice data corresponding to the voice. The storage unit stores the voice data inputted from the input unit. The operation unit is operated by a user. The controller determines a voice activity period in which voice recognition is performed on the voice data stored in the storage unit based on a user's operation on the operation unit. Then, upon detecting a predetermined operation on the operation unit made by the user, the controller adjusts the determined voice activity period.