Vehicle Speech Recognition Gain Control for Startup Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech recognition systems face challenges in setting appropriate input gain, leading to frequent failures immediately after startup, and require complex configurations to learn statistical distributions of audio signals, which can be discouraging for users.
Innovation Solution
A speech recognition system mounted on a vehicle that includes a microphone, a variable-gain input amplifier, a speech recognition unit, a vehicle speed detector, and a gain controller, which uses pre-stored gain control information to set the amplifier gain based on vehicle speed and additional environmental factors like window state and tunnel status, ensuring appropriate gain settings without complex configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If statistical distribution learning or mean amplitude distribution calculation is used to set input gain, then the input gain can be appropriately set according to audio signal characteristics, but the system requires complex configuration and cannot set gain appropriately until statistically significant data is acquired
Solution Approach 1:
The patent pre-calculates and stores the relationship between vehicle speeds and optimal input gain values in a table during system development. During actual operation, the system simply looks up the pre-computed gain value corresponding to the current vehicle speed, eliminating the need for complex statistical learning or amplitude distribution calculation at runtime. This preliminary preparation resolves the contradiction by providing accurate gain setting without requiring complex configuration or extensive data collection.
2Reliability
If statistical distribution learning is used to set input gain, then the gain can be optimized for speech recognition accuracy, but the system experiences frequent failures immediately after startup due to insufficient data
Solution Approach 1:
The optimal input gain values for different vehicle speeds are pre-computed and stored in a lookup table during system development. When the system starts up, it immediately uses the pre-computed gain value corresponding to the current vehicle speed without needing to collect or analyze statistical data. This eliminates the time delay and reliability issues associated with data collection, while still providing optimized gain settings for accurate speech recognition.
3Adaptability or versatility
If complex statistical learning configuration is implemented, then the input gain can be dynamically adjusted according to audio characteristics, but the user experience deteriorates due to frequent failures and complex setup
Solution Approach 1:
The system pre-computes and stores optimal gain values for various vehicle speeds in a lookup table during development. During operation, the system automatically selects the appropriate gain value based on the current vehicle speed without requiring any user configuration or data collection. This maintains the adaptability to different driving conditions while completely eliminating complex setup procedures and frequent failures, thereby improving user ease of use.
Data Source
AI summary
When an instruction to start voice input is received from the user, a gain controller acquires, from a gain table which defines a correspondence between vehicle speed ranges and gains, a gain corresponding to a vehicle speed range including the vehicle speed of a vehicle detected by a vehicle speed detector, and sets the acquired gain as the gain of an input amplifier that amplifies an input audio signal output by a microphone. As a gain corresponding to each vehicle speed range, the gain table records a gain of the input amplifier corresponding, in an experimentally determined frequency distribution of peak values in the vehicle speed range, to a maximum frequency in the range of magnitude of voice output as an input audio signal by the microphone and to be input to a speech recognition engine as voice having a magnitude within the input range of the speech recognition engine.