Vehicle Voice Communication With Motion-Linked Artificial Speech

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing voice communication systems for vehicles struggle to distinguish an artificial voice generated by a machine learning model from a human voice, leading to user misrecognition.

Innovation Solution

A voice communication system that generates an artificial voice similar to a human voice using a machine learning model and adjusts a parameter related to the voice's property (such as pitch) based on a parameter related to the vehicle's motion or driver input, making it easier for users to differentiate the artificial voice from a human voice.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a machine learning model is applied to generate artificial voice similar to human voice, then the voice quality and naturalness are improved, but the user cannot easily distinguish the artificial voice from actual human voice

Engineering Contradiction:
Improvevoice qualityVSAvoiddistinguishability
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent applies parameter changes by modifying voice parameters (pitch, tone, speed) based on vehicle motion parameters (acceleration, deceleration, steering angle). This creates a linkage where voice characteristics dynamically change in response to vehicle movements, making the artificial voice distinguishable from human voice while maintaining naturalness. For example, when the vehicle accelerates, the system adjusts the pitch or tone of the artificial voice to reflect this motion, creating a unique pattern that reveals its artificial origin.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If the artificial voice is made more similar to human voice, then user comfort is improved, but system transparency deteriorates

Engineering Contradiction:
Improveuser comfortVSAvoidsystem transparency
Core Design Contradiction:
Ease of operationVSLoss of information

Solution Approach 1:

The system implements feedback by continuously monitoring vehicle motion parameters and dynamically adjusting voice parameters in response. This creates a closed-loop system where the artificial voice provides feedback about vehicle state through its characteristics. The voice changes in response to acceleration, deceleration, and steering inputs, creating a transparent indication of the system's artificial nature while maintaining user comfort through natural-sounding speech.

Inventive Principle:
Principle #23Feedback

3Stability of the object's composition

If voice parameters are kept constant, then voice stability is improved, but adaptability to vehicle conditions deteriorates

Engineering Contradiction:
Improvevoice stabilityVSAvoidadaptability to vehicle conditions
Core Design Contradiction:
Stability of the object's compositionVSAdaptability or versatility

Solution Approach 1:

The patent applies dynamics by making voice parameters dynamic rather than static. The system continuously adjusts pitch, tone, and speed based on real-time vehicle motion data. When the vehicle accelerates, the voice characteristics change accordingly; when the vehicle decelerates or steers, the voice parameters adapt to reflect these conditions. This dynamic adaptation maintains voice stability within each state while providing overall adaptability across different vehicle conditions.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4495931A1System, method, and program for communicating voice
Publication Date: 2025.01.22 TOYOTA JIDOSHA KK
  • EP4495931A1 patent drawingFigure 1
  • EP4495931A1 patent drawingFigure 2
  • EP4495931A1 patent drawingFigure 3

AI summary

The present disclosure provides a voice communication system (100). The voice communication system (100) comprises one or more processors (11). The one or more processors (11) generate an artificial voice similar to a human voice by a machine learning model. Then, the one or more processors (11) communicate the generated artificial voice to a user of a vehicle (1). The one or more processors (11) acquire a first parameter related to at least one of a motion of the vehicle (1) and an operation amount input by a driver of the vehicle (1) and change a second parameter related to a property of the artificial voice so as to be linked to the first parameter.