Voice Recognition Beamforming Adaptability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional voice recognition technologies using beamforming face challenges in accurately recognizing user input when the beamforming direction does not align with the speaker's direction, leading to decreased recognition rates due to incomplete or missed syllables, especially during the initial tracking time.
Innovation Solution
The proposed solution involves dividing the voice recognition process into two sections: a first section for initial beam direction adjustment and a second section for accurate recognition once the beam is formed, allowing for differentiated processing of voice signals in each phase to improve alignment and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If beamforming is applied for multi-directional voice recognition, then the device can track user direction, but the voice recognition rate decreases during the tracking time when beam direction does not align with speaker direction
Solution Approach 1:
The patent segments the voice recognition process into two distinct sections: a first section that processes voice signals before beamforming convergence, and a second section that processes voice signals after beamforming convergence. This segmentation allows different recognition strategies to be applied at different stages, resolving the contradiction between multi-directional adaptability and recognition reliability during the transition period.
Solution Approach 2:
The patent performs preliminary voice recognition processing in the first section before the beamforming operation is fully converged. By conducting recognition operations in advance during the tracking time when beam direction may not align with speaker direction, the system ensures that voice input is not lost during the transition period, thereby maintaining recognition reliability while preserving multi-directional capability.
2Adaptability or versatility
If beamforming direction tracking is implemented, then the device can adapt to speaker position, but syllables may be cut-off or missed during the predetermined tracking time
Solution Approach 1:
The patent divides the voice signal processing into two temporal segments: a first section covering the period before beamforming convergence where voice signals may be incomplete, and a second section covering the period after convergence where complete recognition occurs. This segmentation prevents information loss by ensuring that partial signals during tracking are still processed through recognition operations.
Solution Approach 2:
The patent maintains continuous voice recognition processing across both sections, ensuring that the useful action of recognizing voice input continues uninterrupted. By performing recognition operations in both the pre-convergence and post-convergence sections, the system ensures that no voice information is lost during the beam direction tracking transition, maintaining signal completeness.
3Device complexity
If conventional voice recognition is used without beamforming training, then the system is simpler to implement, but the voice recognition rate is reduced
Solution Approach 1:
The patent segments the recognition system into two operational modes corresponding to the two sections: a first recognition mode for the pre-convergence period and a second recognition mode for the post-convergence period. This segmentation allows the system to use simpler processing in the first section while applying more sophisticated beamforming-based recognition in the second section, achieving improved recognition rates without requiring complete system complexity.
Data Source
AI summary
The present disclosure relates to an electronic device and a method for operating same, which can divide a recognition section for voice recognition, process different types of voice recognition corresponding to the divided recognition sections, and improve a voice recognition rate. According to an embodiment, a method for recognizing a voice by an electronic device may include: detecting a voice input; determining the direction of the voice and a beamforming direction. Voice recognition is based on the voice when the direction of the voice and the beamforming direction correspond to each other. Different types of voice recognition may be performed for voices input in a first section and a second section, based on divided types of voice input sections, when the direction of the voice and the beamforming direction do not correspond to each other.