Voice Gesture Recognition Fusion for Vehicle Command Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional human machine interfaces (HMIs) in vehicles are limited in recognizing user commands based on voice or visual signals, lacking robustness in identifying user requests.
Innovation Solution
A system and method that combines voice recognition and gesture recognition, using confidence scores from both modalities to immediately execute commands, request confirmation, or determine unrecognized inputs, by integrating voice and gesture recognition modules within a vehicle's communication system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional voice recognition or gesture recognition is used separately, then the system structure is simple, but the command recognition accuracy is insufficient
Solution Approach 1:
The patent combines voice recognition module and gesture recognition module into a unified HMI system, where both modules process their respective inputs and the results are integrated through confidence score comparison to determine the final command, thereby improving recognition accuracy while maintaining manageable system complexity through modular integration
2Reliability
If multiple recognition modalities are integrated, then the robustness of command identification is improved, but the processing complexity increases
Solution Approach 1:
The system implements a feedback mechanism where confidence scores from both voice and gesture recognition are continuously evaluated against threshold criteria. The processor adjusts its decision-making based on the combined confidence levels, requesting clarification or switching modalities when confidence is insufficient, thereby improving reliability through adaptive feedback processing
3Measurement precision
If confidence threshold evaluation is performed for both voice and gesture, then the command execution accuracy is improved, but the response time increases
Solution Approach 1:
The system performs preliminary confidence evaluation on both voice and gesture inputs simultaneously as they are received, rather than sequentially processing one modality after the other. This parallel preliminary assessment allows the system to quickly determine whether the combined confidence threshold is met, reducing response time while maintaining accurate command execution
Data Source
AI summary
A system and method for combining two separate types of human machine interfaces, e.g., a voice signal and a gesture signal, performing voice recognition to a voice signal and gesture recognition to the gesture signal. Based on a confidence determination using the voice recognition result and the gesture recognition result the system can, for example, immediately perform the command/request, request confirmation of the command/request or determine that the command/request was not identified.