Methods, apparatus, and articles of manufacture to generate voices for artificial speech based on an identifier represented by frequency dependent bits

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing voice-based interactive devices have limited voice options, making it difficult for users to distinguish between devices, and current systems require complex configuration and maintenance, failing to provide unique voices for each device.

Innovation Solution

The use of unique device-specific identifiers, such as serial numbers or MAC addresses, to personalize artificial speech output, generating distinct voices for each device through audio synthesizers and mixers, allowing devices to produce audibly different voices even when using the same base voice.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a limited number of voices are provided in voice-based interactive devices, then device complexity is reduced, but users cannot distinguish between devices and voice personalization is lost

Engineering Contradiction:
Improvevoice configuration complexityVSAvoidvoice distinction capability
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent applies local quality by modifying specific acoustic parameters (pitch, tone, timbre) of the base voice to create device-specific voice characteristics. Each device receives a customized voice variant based on its unique identifier, allowing voice differentiation without requiring multiple complete voice sets or complex configuration systems.

Inventive Principle:
Principle #3Local quality

2Loss of information

If multiple voice options are provided for each device, then voice differentiation is improved, but configuration complexity and maintenance burden increase

Engineering Contradiction:
Improvevoice distinction capabilityVSAvoidvoice configuration complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The system implements self-service by automatically generating device-specific voice characteristics using the device's unique identifier. The voice personalization process occurs automatically during device initialization without requiring user configuration or manual setup, eliminating the maintenance burden associated with managing multiple voice options while preserving voice differentiation capability.

Inventive Principle:
Principle #25Self-service

3Loss of information

If device-specific voice personalization is implemented, then voice distinction between devices is achieved, but system complexity and maintenance costs increase

Engineering Contradiction:
Improvevoice distinction capabilityVSAvoidsystem implementation complexity
Core Design Contradiction:
Loss of informationVSEase of manufacture

Solution Approach 1:

The patent implements parameter changes by modifying acoustic parameters (pitch, tone, timbre) of a single base voice based on the device's unique identifier. This approach achieves voice distinction by adjusting voice parameters rather than implementing entirely separate voice systems, significantly reducing system implementation complexity and maintenance costs while preserving full voice differentiation capability.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS10468013B2Methods, apparatus, and articles of manufacture to generate voices for artificial speech based on an identifier represented by frequency dependent bits
Publication Date: 2019.11.05 INTEL CORP
  • US10468013B2 patent drawing
  • US10468013B2 patent drawing
  • US10468013B2 patent drawing

AI summary

Methods, apparatus, and articles of manufacture to generate voices for artificial speech are disclosed. An example apparatus includes a component storing an identifier, the identifier uniquely identifying the apparatus from a plurality of apparatus, an artificial speech generator to generate a first artificial speech signal representing text, the first artificial speech signal generated based on the identifier which is represented by frequency dependent bits by assigning specific bits to specific frequency bands, the first artificial speech signal audibly different from artificial speech signals generated by respective ones of the plurality of apparatus for the text, and an output device to output an audible signal representing the first artificial speech signal.