Speech Recognition Device Command Verification Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional speech recognition devices do not differentiate the importance of the content when deciding whether to perform a confirmation process, leading to unnecessary user interactions.
Innovation Solution
A speech recognition device that includes a speech reception unit, a speech recognition unit, a condition storage unit, and a determination unit to assess the reliability of recognized commands and determine if a confirmation process is necessary based on pre-defined conditions associated with each command, thereby reducing unnecessary user interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a confirmation process is executed for all speech recognition results regardless of content importance, then the reliability of information can be ensured, but the number of user operations increases and user-friendliness deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the confirmation process based on the specific command type. Different commands have different importance levels, and the confirmation threshold is adjusted locally for each command type. High-importance commands (e.g., safety-critical operations) require confirmation even with high reliability, while low-importance commands can proceed without confirmation if reliability exceeds their specific threshold.
Solution Approach 2:
The patent changes the parameter of confirmation threshold dynamically based on command importance. Instead of using a fixed confirmation policy, the system adjusts the reliability threshold parameter according to the specific command being recognized. This allows the system to maintain high reliability for critical commands while reducing unnecessary confirmations for non-critical commands.
2Reliability
If a confirmation process is executed for all speech recognition results, then errors can be caught, but the processing time and operational complexity increase
Solution Approach 1:
The patent implements local quality by applying different confirmation requirements to different command types based on their importance and potential consequences. Critical commands that could cause safety issues or significant errors maintain strict confirmation requirements, while routine or low-impact commands use relaxed or no confirmation, thereby reducing overall processing time.
Solution Approach 2:
The patent applies partial action by selectively applying confirmation processes only to commands where it is truly necessary. Instead of universally applying confirmation to all commands, the system identifies and applies confirmation only to high-importance commands, achieving sufficient error prevention without the time cost of universal confirmation.
3Device complexity
If the confirmation process is determined solely by reliability threshold without considering command importance, then the decision process is simple, but unnecessary confirmations occur for low-importance commands
Solution Approach 1:
The patent applies local quality by assigning different importance levels and confirmation thresholds to different command types. The determination process remains relatively simple through automated classification, but it now considers the specific command context locally to decide whether confirmation is necessary, reducing unnecessary user operations for low-importance commands.
Solution Approach 2:
The patent implements preliminary action by pre-classifying commands into importance categories and pre-setting appropriate confirmation thresholds for each category. This preliminary classification and threshold assignment automates the decision-making process, making it both simple to execute and sensitive to command-specific requirements.
Data Source
AI summary
This speech recognition device comprises: a speech acceptance unit for accepting speech information representing one command among a plurality of commands; a speech recognition unit for performing speech recognition of the one command on the basis of the speech information accepted by the speech acceptance unit and for computing a reliability score for the recognition result of the one command; a condition storage unit for storing a plurality of conditions used in determining whether to execute a verification process for a recognition result, in association with a plurality of respective commands; a determination unit for determining whether to execute a verification process on the basis of one condition among a plurality of conditions stored in the condition storage unit and the reliability score computed by the speech recognition unit; and an output unit for outputting the recognition result without executing the verification process in a case where the determination unit determined to not execute the verification process.