Voice Recognition Threshold Adjustment by Sound Direction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice recognition systems face challenges in adapting to changing environments, particularly in noisy conditions, due to the use of a single threshold value for wake-up word recognition, leading to increased misrecognition rates and user inconvenience.
Innovation Solution
An electronic apparatus that identifies the direction of sound reception and adjusts threshold values based on similarity between user utterance and wake-up word, using multiple threshold values to improve recognition accuracy and provide guidance for noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a high threshold value is set to reduce misrecognition rate in noisy environment, then misrecognition rate is reduced, but voice recognition becomes difficult to use and user convenience deteriorates
Solution Approach 1:
The patent applies dynamics by making the threshold value adjustable rather than fixed. The threshold is dynamically changed based on the detected receiving direction of sound and the similarity between detected sound and wake-up word. When noise is detected from specific directions, the threshold is adjusted to maintain both accuracy and usability.
Solution Approach 2:
The patent changes the threshold parameter based on environmental conditions. Multiple threshold values are defined corresponding to different receiving directions, and the appropriate threshold is selected or adjusted based on the detected sound direction and similarity metrics, allowing the system to adapt to varying acoustic environments.
2Device complexity
If a single threshold value is used for wake-up word recognition, then device complexity is reduced, but adaptability to changing environments deteriorates
Solution Approach 1:
The patent segments the single threshold value into multiple threshold values corresponding to different receiving directions. Each direction has its own threshold (first threshold value for first direction, second threshold value for second direction, etc.), allowing the system to handle different acoustic environments separately while maintaining manageable complexity through automated selection.
Solution Approach 2:
The system dynamically selects or adjusts the appropriate threshold based on the detected receiving direction. Rather than using a static single threshold, the system adapts the threshold parameter in real-time based on environmental conditions, improving versatility without requiring complex manual configuration.
3Ease of manufacture
If the threshold value is changed collectively for all directions, then implementation is simplified, but recognition rate deteriorates in specific directions
Solution Approach 1:
The patent applies local quality by assigning different threshold values to different receiving directions. Each direction receives a threshold optimized for its specific characteristics (e.g., first threshold value for first direction, second threshold value for second direction). This allows the system to optimize recognition performance for each direction independently while maintaining a unified overall structure.
Data Source
AI summary
An electronic apparatus includes: a memory storing a first threshold value and a second threshold value corresponding to a receiving direction of a wake-up word, a sound receiver comprising sound receiving circuitry, and a processor configured to: identify a receiving direction of the sound based on a sound received through the sound receiver, based on a similarity between sound data obtained in response to the received sound and the wake-up word being greater than or equal to the first threshold value corresponding to the identified receiving direction, perform voice recognition for a subsequent sound received through the sound receiver, and based on the similarity being less than the first threshold value and greater than or equal to the second threshold value, change the first threshold value.