Voice to Text Biasing via Third-Party Context

Solution Overview

Problem

Existing voice-enabled electronic devices face challenges in accurately converting voice input to text, especially in context-dependent scenarios, leading to potential inaccuracies and increased computational resource usage.

Innovation Solution

The implementation of a voice to text engine within a local agent that dynamically biases voice to text conversions based on contextual parameters provided by a third-party agent, enhancing the accuracy and robustness of the conversion process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If standard voice to text conversion is used without contextual biasing, then the conversion process is simpler and faster, but the accuracy of the text representation decreases

Engineering Contradiction:

Improvevoice to text conversion accuracyVSAvoidconversion process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary actions by obtaining contextual parameters from the third-party agent before the voice to text conversion occurs. These contextual parameters are used to bias the conversion process, preparing the system in advance to handle the conversion more accurately by having relevant context ready beforehand.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from the third-party agent in the form of contextual parameters that indicate potential features of further voice input. This feedback loop allows the voice to text engine to continuously adapt and improve conversion accuracy based on the agent's expectations and the ongoing dialog context.

Inventive Principle:

Principle #23Feedback

2Measurement precision

If contextual parameters from third-party agent are used to bias voice to text conversion, then conversion accuracy improves, but computational resource usage increases

Engineering Contradiction:

Improvevoice to text conversion accuracyVSAvoidcomputational resource usage

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system applies partial action by using contextual parameters selectively to bias only certain aspects of the voice to text conversion process. Rather than completely reprocessing the voice input, the system partially modifies the conversion based on contextual cues, reducing the overall computational burden while still improving accuracy.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The system changes parameters of the voice to text model dynamically based on contextual parameters from the third-party agent. By adjusting model parameters rather than retraining or replacing the entire model, the system achieves improved accuracy with minimal additional computational overhead.

Inventive Principle:

Principle #35Parameter changes

3Productivity

If voice to text conversion is performed without contextual biasing, then processing is faster, but inaccuracies in voice input representation increase

Engineering Contradiction:

Improveprocessing speedVSAvoidvoice input representation accuracy

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary actions by obtaining contextual parameters from the third-party agent before the voice to text conversion occurs. These contextual parameters are used to bias the conversion process, preparing the system in advance to handle the conversion more accurately by having relevant context ready beforehand.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from the third-party agent in the form of contextual parameters that indicate potential features of further voice input. This feedback loop allows the voice to text engine to continuously adapt and improve conversion accuracy based on the agent's expectations and the ongoing dialog context.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12217759B2Voice to text conversion based on third-party agent content

Publication Date: 2025.02.04 GOOGLE LLC

AI summary

Implementations relate to dynamically, and in a context-sensitive manner, biasing voice to text conversion. In some implementations, the biasing of voice to text conversions is performed by a voice to text engine of a local agent, and the biasing is based at least in part on content provided to the local agent by a third-party (3P) agent that is in network communication with the local agent. In some of those implementations, the content includes contextual parameters that are provided by the 3P agent in combination with responsive content generated by the 3P agent during a dialog that: is between the 3P agent, and a user of a voice-enabled electronic device; and is facilitated by the local agent. The contextual parameters indicate potential feature(s) of further voice input that is to be provided in response to the responsive content generated by the 3P agent.