Accent-Resilient Speech Transmission via Multi-Converter Error Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Communications in the airline and avionics industries face challenges due to accent-induced anomalies, particularly in non-native languages, leading to inaccuracies and increased effort required for understanding instructions, as existing solutions like voice-to-text applications may result in inaccuracies.

Innovation Solution

A speech transmission system that performs multiple independent voice-to-text conversions, compares results for error detection, and uses error codes to replace unrecognized words, with a text-to-voice conversion to generate a corrected audio stream, ensuring high accuracy by using different conversion algorithms and error code substitution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a single voice-to-text conversion is used, then the system complexity is low, but the accuracy of speech transmission deteriorates due to accent-induced errors

Engineering Contradiction:
Improveaccuracy of speech transmissionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system divides the speech-to-text conversion process into multiple independent conversion channels, each processing the same audio input through different algorithms. This segmentation allows the system to compare results from multiple converters and identify errors, thereby improving accuracy while managing complexity through modular design

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system implements a feedback mechanism where the outputs of multiple voice-to-text converters are compared against each other. Discrepancies detected through this feedback loop trigger error correction processes, allowing the system to identify and correct accent-induced errors dynamically

Inventive Principle:
Principle #23Feedback

2Measurement precision

If multiple independent voice-to-text conversions are performed, then the accuracy of speech transmission is improved, but the processing time increases

Engineering Contradiction:
Improveaccuracy of speech transmissionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary error detection by comparing the outputs of multiple voice-to-text converters before final text-to-voice conversion. This preliminary action identifies discrepancies early in the process, allowing for targeted corrections rather than reprocessing entire sequences, thus reducing overall processing time

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses multiple conversion algorithms simultaneously (excessive action) but only processes the comparison and correction of discrepancies (partial action). This approach leverages parallel processing to maintain speed while achieving higher accuracy through selective error correction

Inventive Principle:
Principle #16Partial or excessive action

3Reliability

If voice-to-text applications are used for correction, then communication accuracy may improve, but the system becomes more complex and may introduce additional errors

Engineering Contradiction:
Improvecommunication accuracyVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system introduces an intermediary comparison mechanism that mediates between multiple voice-to-text converters. This intermediary layer compares outputs, identifies discrepancies, and selects the most accurate interpretation, adding reliability while managing complexity through a structured mediation process rather than uncontrolled system complexity

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP2770501B1System and method for correcting accent induced speech transmission problems
Publication Date: 2017.05.31 HONEYWELL INTERNATIONAL INC
  • EP2770501B1 patent drawingFigure 1
  • EP2770501B1 patent drawingFigure 2
  • EP2770501B1 patent drawingFigure 3

AI summary

A system and method is provided for detecting errors in a speech transmission system. A first audio stream is comprised of a plurality of words, upon which a plurality of independent voice-to-text conversions are performed. If it is determined that at least one of the plurality of independent voice-to-text conversions is error free, a text-to-voice conversion of the at least one error-free voice-to-text conversion is performed to create a second audio stream.