Voice Input Correction via Confidence-Based Self-Service
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current voice input interfaces often misinterpret user inputs, requiring manual correction by users, which is intrusive and not intuitive, especially when the system fails to identify low-confidence words or misinterpretations.
Innovation Solution
A method and system that allow users to provide corrective voice input to correct mistakes seamlessly within the voice interface, using supplemental voice input that mimics natural communication, where the user can speak corrections without pre-defined voice commands, enabling quick and intuitive error correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If users manually correct voice input errors by deleting and re-inputting text, then correction accuracy is improved, but operation time and user convenience deteriorate
Solution Approach 1:
The system automatically identifies and corrects voice input errors without requiring user intervention. The speech recognition system self-corrects by detecting confidence levels of recognized words and automatically replacing misrecognized words with correct alternatives, eliminating the need for manual deletion and re-inputting by the user.
Solution Approach 2:
The system provides feedback to the user by highlighting potential errors in the recognized text and offering correction options. This feedback mechanism allows the system to identify low-confidence recognitions and present them for automatic correction, reducing the time users would otherwise spend manually correcting errors.
2Measurement precision
If users select low-confidence text for correction from a drop-down list, then correction precision is improved, but ease of operation deteriorates
Solution Approach 1:
The system automatically identifies and corrects errors without requiring users to manually select from drop-down lists. The speech recognition system self-corrects by detecting confidence levels and automatically replacing misrecognized words, eliminating the need for users to manually navigate and select corrections from lists.
Solution Approach 2:
The system changes the parameter of correction approach from manual selection to automatic identification. By monitoring confidence levels of recognized words and automatically replacing low-confidence recognitions, the system transforms the correction process from a manual interaction requiring user selection to an automated process that maintains precision without sacrificing ease of operation.
3Reliability
If the voice input interface provides manual correction options, then reliability is improved, but device complexity increases
Solution Approach 1:
The system uses automatic self-correction mechanisms that monitor confidence levels of speech recognition output and automatically replace misrecognized words. This self-service approach maintains high reliability by catching errors automatically without requiring complex manual correction interfaces, thereby improving reliability while avoiding increased device complexity.
Data Source
AI summary
An embodiment provides a method, including: accepting, at an audio receiver of an information handling device, voice input of a user; interpreting, using a processor, the voice input; thereafter receiving, at the audio receiver, repeated voice input of the user; identifying a correction using the repeated voice input; and correcting, using the processor, the voice input using the repeated voice input, wherein the corrective voice input does not include a predetermined voice command. Other aspects are described and claimed.