Speech Recognition Threshold Adjustment for Expected Responses
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition systems often reject valid speech inputs due to confidence factors falling below the acceptance threshold, leading to reduced productivity and efficiency, especially in environments where expected responses are known.
Innovation Solution
Modifying the acceptance threshold based on knowledge of expected responses, allowing speech recognition systems to accept inputs with lower confidence factors when they match predetermined expected responses, thereby reducing unnecessary repetition and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the speech recognition system uses a fixed acceptance threshold to ensure recognition accuracy, then the reliability of speech recognition is improved, but the productivity deteriorates due to frequent rejections of valid inputs requiring repetition
Solution Approach 1:
The patent applies dynamics by making the acceptance threshold adjustable rather than fixed. The threshold is dynamically modified based on whether the current speech input matches an expected response. When a match is detected, the threshold is lowered to accept inputs with lower confidence factors, thereby improving productivity without sacrificing overall recognition reliability.
Solution Approach 2:
The patent changes the parameter of the acceptance threshold based on the context of expected responses. By modifying this critical parameter conditionally, the system achieves both high reliability (through maintained accuracy for non-expected inputs) and high productivity (through increased acceptance of expected inputs even with lower confidence).
2Productivity
If the speech recognition system lowers the acceptance threshold to reduce repetition, then the productivity is improved, but the reliability deteriorates due to increased acceptance of incorrect inputs
Solution Approach 1:
The patent applies local quality by applying different acceptance threshold levels to different types of speech inputs. Expected responses (which are known in advance) receive a lower, more permissive threshold, while other inputs maintain the original higher threshold. This localized differentiation allows productivity improvement for expected inputs without compromising reliability for unexpected inputs.
Solution Approach 2:
The system performs preliminary action by identifying expected responses before the speech recognition decision is made. By knowing in advance what responses are expected, the system can prepare to apply a modified threshold specifically for those cases, ensuring that productivity improvements are targeted and do not broadly compromise recognition accuracy.
3Measurement precision
If the speech recognition system requires high confidence factors for acceptance, then the measurement precision is improved, but the loss of time increases due to repeated speech inputs
Solution Approach 1:
The patent makes the confidence factor requirement dynamic rather than static. For expected responses, the system dynamically reduces the confidence factor requirement, allowing acceptance of inputs that would otherwise be rejected due to lower confidence scores. This dynamic adjustment eliminates unnecessary repetition and reduces time loss while maintaining precision for non-expected inputs.
Data Source
AI summary
A speech recognition system receives and analyzes speech input from a user in order to recognize and accept a response from the user. Under certain conditions, information about the response expected from the user may be available. In these situations, the available information about the expected response is used to modify the behavior of the speech recognition system by taking this information into account. The modified behavior of the speech recognition system comprises adjusting the rejection threshold when speech input matches the predetermined expected response.