Voice Assistant Text Feedback Confidence Prioritization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face challenges in selecting the most appropriate option from a crowded vehicle display, particularly during tasks like navigation or phone calls, due to the time-consuming nature of focusing on the screen while driving, as existing speech recognition systems do not effectively prioritize results based on user habits or context.
Innovation Solution
The system consolidates speech recognition results by incorporating a priori information on frequent contacts, visited points of interest, and commonly used commands, and adjusts font sizes and voice feedback volumes based on confidence scores to prioritize likely results, enhancing user convenience by accentuating the most probable options.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the system displays all speech recognition results with equal visual prominence, then the user can see all options, but the user spends more time selecting the most appropriate option
Solution Approach 1:
The patent applies local quality by differentiating the visual presentation of speech recognition results based on their confidence scores. High-confidence results are displayed with enhanced visual characteristics (larger font size, brighter color, bold weighting) while lower-confidence results receive standard or reduced visual emphasis. This selective differentiation allows users to quickly identify the most likely correct option without losing visibility of alternative interpretations.
2Adaptability or versatility
If the system presents multiple speech recognition hypotheses with equal emphasis, then all options are visible, but the cognitive load on the user increases
Solution Approach 1:
The system changes visual parameters (font size, color intensity, text weighting) based on the confidence score parameter associated with each speech recognition hypothesis. By dynamically adjusting these visual parameters according to the likelihood of each interpretation, the system maintains adaptability to present all possible options while reducing cognitive load through hierarchical visual organization that guides user attention to the most probable correct answer.
Data Source
AI summary
A method for text feedback includes: receiving, by a controller, an utterance from a user; determining, by an automatic speech recognition engine of the controller, a plurality of speech recognition results based on the utterance from the user, wherein the speech recognition results include probable commands; determining, by the automatic speech recognition engine of the controller, a plurality of confidence scores for each of the plurality of speech recognition results; determining, by the controller, a text characteristic for each of the plurality of probable commands as a function of the confidence scores for each of the plurality of speech recognition results; and commanding, by the controller, a display to show text corresponding to each of the plurality of probable commands with the text characteristic determined by the controller.