Vehicle Speech Recognition Using Spatial Spectrum Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech recognition systems in vehicles face challenges in accurately distinguishing speech from noise generated by vehicle states such as speed, door operations, and wiper activity, leading to degraded recognition accuracy.
Innovation Solution
A speech recognition apparatus that includes a sound collection unit, sound source localization unit, speech zone determination unit, and speech recognition unit, which uses spatial spectrum analysis and threshold values based on vehicle states to isolate and prioritize speech signals from drivers and passengers, reducing noise interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If speech recognition is performed using sound signals collected in a vehicle, then speech recognition functionality is provided, but noise generated by vehicle states (speed, door operations, wiper activity) is recognized as speech, degrading recognition accuracy
Solution Approach 1:
The patent segments the sound signal processing into distinct functional modules: sound collection unit, sound source localization unit, speech zone determination unit, and speech recognition unit. This segmentation allows each unit to perform its specific function optimally, with the speech zone determination unit specifically designed to filter noise based on vehicle state information.
Solution Approach 2:
The patent performs preliminary actions by determining speech zones based on vehicle state information before actual speech recognition occurs. The system pre-establishes which spatial zones are likely to contain valid speech signals by considering vehicle speed, door operations, and wiper activity, thereby filtering out noise sources before they can interfere with recognition.
2Adaptability or versatility
If a fixed threshold value is used for speech zone determination, then the system structure is simple, but it cannot adapt to different vehicle states, reducing speech recognition accuracy
Solution Approach 1:
The patent implements dynamic threshold values that automatically adjust based on vehicle state. Instead of using a fixed threshold, the system changes the speech zone determination threshold according to vehicle speed, door operation status, and wiper activity, allowing the recognition system to adapt to varying noise conditions without manual intervention.
Solution Approach 2:
The system incorporates feedback mechanisms where vehicle state information (speed, door position, wiper status) is continuously monitored and fed back to the speech zone determination unit. This feedback loop enables real-time adjustment of speech zones and threshold values to match current vehicle conditions, improving adaptability while maintaining automated operation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively reduces noise-related errors in speech recognition by identifying and prioritizing speech zones based on vehicle states, enhancing recognition accuracy for both drivers and passengers.
Implementation Method 1
a sound source localization unit that calculates a spatial spectrum from the sound signal that is collected by the sound collection unit and uses the calculated spatial spectrum to perform sound source localization
Data Source
AI summary
A speech recognition apparatus includes: a sound collection unit that collects a sound signal; a sound source localization unit that calculates a spatial spectrum from the sound signal that is collected by the sound collection unit and uses the calculated spatial spectrum to perform sound source localization; a speech zone determination unit that determines a zone in which a power of the spatial spectrum that is calculated by the sound source localization unit exceeds a predetermined threshold value based on a vehicle state; and a speech recognition unit that performs speech recognition with respect to a sound signal of the zone determined by the speech zone determination unit.