Voice Assistant Routing to Specialized Apps for Recognition Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice assistants face challenges in accurately and efficiently executing user instructions due to their broad vocabulary lexicons and general voice recognition algorithms, which can lead to inaccurate task completion when specific functions are required, as they often rely on passing recognition results to application programs that are not specialized in those functions.
Innovation Solution
A voice function control method where a voice assistant determines and passes the user's input voice to a specialized application program that is configured to implement specific functions, allowing the application program to recognize and execute the user's intent more accurately and quickly, leveraging its optimized voice recognition capabilities for that function type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a voice assistant uses a general voice recognition algorithm with broad vocabulary lexicon to handle multiple function types, then the adaptability and versatility are improved, but the voice recognition accuracy deteriorates
Solution Approach 1:
The patent segments the voice recognition task by dividing it into two parts: a voice assistant layer that handles general intent recognition and application selection, and specialized application program layers that handle function-specific voice recognition. This segmentation allows each component to focus on its strength, resolving the contradiction between broad adaptability and specialized accuracy.
Solution Approach 2:
The voice assistant acts as an intermediary between the user and multiple application programs. It receives the input voice, performs initial processing to determine which application program is appropriate, and then passes the voice to the selected application program for specialized recognition and execution. This intermediary role enables the system to maintain both broad versatility and specialized accuracy.
2Device complexity
If a voice assistant passes recognition results to application programs for task completion, then the device complexity is reduced, but the task completion accuracy deteriorates
Solution Approach 1:
Instead of the conventional approach where application programs pass recognition results to a central voice assistant, this patent inverts the flow: the voice assistant passes the input voice to the selected application program, which then performs its own recognition and returns results. This inversion allows application programs to leverage their specialized recognition capabilities while maintaining a simple overall system architecture.
Solution Approach 2:
Application programs perform their own voice recognition using their specialized algorithms and lexicons, rather than relying on the voice assistant's general recognition. This self-service approach enables each application program to optimize its recognition accuracy for its specific function domain while keeping the system architecture simple and modular.
3Ease of operation
If a voice assistant uses a common voice recognition algorithm for all function types, then the ease of operation is improved, but the productivity deteriorates
Solution Approach 1:
The system dynamically adapts its recognition approach based on the selected application program. While the user interface remains consistent and simple (voice assistant receives all input uniformly), the underlying recognition process dynamically switches to the specialized algorithm of the relevant application program, thereby improving task completion speed without compromising ease of operation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present application provides a voice function control method. The voice function control method is applied to a voice assistant of a terminal, and includes: determining an associated application program based on a result of recognizing an input voice of a user, where the associated application program is configured to implement a function to be used by the user; and passing the input voice of the user to the associated application program, so that the associated application program recognizes the input voice of the user, and implements the function based on a recognition result. According to the technical solutions of the present application, the function required by the user can be more accurately and quickly completed, thereby improving voice function control efficiency.