Voice Processing Apparatus Confirmation Question Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice interactive systems face challenges in accurately and efficiently acquiring the intention of a user's utterance, often requiring repeated attempts due to errors in voice recognition, leading to prolonged interaction times.
Innovation Solution
A voice processing apparatus and method that includes a sound input unit, voice recognition unit, intention understanding unit, and question unit, which asks confirmation questions and adjusts question content based on priority, allowing users to select candidates from displayed options, thereby improving recognition accuracy and reducing interaction time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voice recognition is performed on user utterance, then the system can acquire user input, but recognition errors occur requiring repeated attempts
Solution Approach 1:
The system asks confirmation questions to the user based on the understood result, creating a feedback loop that verifies recognition accuracy. The question unit generates confirmation questions about the understood user intention, and the user's responses provide feedback to correct any recognition errors, thereby improving accuracy while reducing repeated attempts.
Solution Approach 2:
The system performs preliminary voice recognition to derive an understood result before asking confirmation questions. This preliminary action allows the system to proactively identify potential recognition issues and address them through targeted confirmation questions, reducing the need for repeated full recognition attempts.
2Measurement precision
If the system asks multiple questions to clarify user intention, then recognition accuracy improves, but user burden increases
Solution Approach 1:
The question unit segments the clarification process into targeted confirmation questions based on the understood result. Instead of asking multiple generic questions, the system asks specific confirmation questions about particular aspects of the understood user intention, reducing the total number of questions needed while maintaining accuracy.
Solution Approach 2:
The system asks confirmation questions selectively based on the confidence level and specific uncertainties in the understood result. Rather than asking all possible questions, it asks only the necessary partial set of questions to resolve ambiguities, reducing user burden while achieving sufficient accuracy.
Data Source
AI summary
A voice processing apparatus includes: a sound input unit configured to acquire an audio signal; a voice recognition unit configured to perform voice recognition on the audio signal acquired by the sound input unit; an intention understanding unit configured to understand a user's intention on the basis of a recognition result recognized by the voice recognition unit; and a question unit configured to question the user on the basis of an understood result understood by the intention understanding unit. The question unit changes question content for the user according to the understood result and a predetermined priority.