Vehicle Guidance Speech Correction for Ambient Noise and Speech Heads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech correction apparatuses for car-mounted navigation systems often result in excessive correction of the head portion of guidance speech, leading to an unnatural waveform due to incorrect detection of sound pressure levels and initial filter coefficient settings, which affects the clarity and audibility of navigation instructions.
Innovation Solution
A speech correction apparatus that includes a microphone, a signal separating unit, a speech correcting unit, and a speech-head correcting unit, which adjusts the average power calculation for the head portion of speech by using different time lengths for ambient noise and speech, and sets initial filter coefficients to prevent excessive correction, ensuring accurate sound pressure level adjustment and maintaining clarity across varying noise levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the sound pressure level is detected by averaging the power over a predetermined time period, then the gain can be determined for amplifying the guidance speech, but the sound pressure level at the head portion is wrongly detected at a lower value than the actual value, causing excessive correction
Solution Approach 1:
The patent divides the speech signal processing into two separate paths: one for detecting the sound pressure level of the entire guidance speech (including head portion) by averaging power over a predetermined time period, and another for detecting the sound pressure level of only the non-head portion. This segmentation allows the system to exclude the head portion from excessive correction while still accounting for it in the overall gain determination.
Solution Approach 2:
The patent applies different correction strategies to different parts of the speech signal. The head portion (first predetermined time period) is detected separately and excluded from the averaging calculation that determines correction gain, while the non-head portion undergoes normal gain correction based on ambient noise levels. This local differentiation prevents the unnatural pulse-like waveform that would result from uniform correction across the entire speech signal.
2Adaptability or versatility
If the filter coefficient is updated from 0 in the initial state, then the adaptive filter can adapt to the acoustic environment, but the power of guidance speech estimated by using the adaptive filter is much smaller than the actual value, causing excessive correction of the head portion
Solution Approach 1:
The patent implements a preliminary detection phase where the head portion of the speech (during the first predetermined time period after speech detection) is identified and excluded from the power averaging calculation. This preliminary action occurs before the adaptive filter has had sufficient time to converge and provide accurate power estimates, thereby preventing excessive correction that would result from using inaccurate initial estimates.
Solution Approach 2:
The patent introduces dynamic time-period-based processing where the correction gain calculation changes based on the temporal position of the speech signal. During the initial head portion (first predetermined time period), a different processing mode is applied compared to the steady-state speech portion. This dynamic approach allows the system to adapt to the transient behavior of the adaptive filter during initialization while maintaining accurate correction during stable operation.
Data Source
AI summary
A speech correction apparatus includes a speaker for generating guidance speech; a microphone set at a hearing position; an acoustic-characteristic setting unit for separating ambient noise from the guidance speech at the hearing position; an operating unit; a speech correcting filter for correcting the sound pressure level of the guidance speech generated by the speaker based on the average power of the guidance speech and the average power of the ambient noise which are separated; a loudness-compensating-gain calculating unit; and a speech-head correcting unit for correcting the average power of the guidance speech corresponding to the speech head at the border between a silent state and a speech state of the guidance speech.