In-Vehicle Audio Speech Removal for Natural Sound
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing in-vehicle audio systems struggle to distinguish and correct for speech sounds within noise, leading to unnatural sound adjustments due to sudden changes in speech frequency and volume, which affects the quality of audio playback.
Innovation Solution
An audio apparatus with a noise-extracting unit that includes a speech-removing unit, using pre-defined noise spectrum data to differentiate and remove speech signals from noise, allowing for a weighting factor calculation to correct audio signals based on the cleaned noise, thereby producing a more natural sound quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If noise correction is performed based on noise level including speech sound, then audio volume is adjusted to compensate for background noise, but the sudden changes in speech frequency and volume cause large correction changes resulting in unnatural sound
Solution Approach 1:
The patent extracts and removes speech signals from the noise signal using spectral analysis. The speech-removing unit identifies frequency bands where speech sounds occur and eliminates these components, leaving only genuine background noise for volume correction. This prevents speech-induced sudden volume changes from causing unnatural audio corrections.
Solution Approach 2:
The patent segments the noise signal into different frequency bands and analyzes the spectral characteristics to distinguish between speech sounds and actual background noise. By processing different frequency components separately, the system can identify and remove speech while preserving genuine noise characteristics for accurate volume adjustment.
2Adaptability or versatility
If speech sound is included in noise signal for correction, then volume adjustment responds to all sounds, but speech causes sudden frequency and volume changes that make correction unnatural
Solution Approach 1:
The patent extracts and removes speech signals from the noise signal using spectral analysis. The speech-removing unit identifies frequency bands where speech sounds occur and eliminates these components, leaving only genuine background noise for volume correction. This prevents speech-induced sudden volume changes from causing unnatural audio corrections.
Solution Approach 2:
The patent performs preliminary spectral analysis and speech detection before executing volume correction. By preparing noise spectrum data and identifying speech components in advance, the system can exclude speech from the correction calculation, ensuring stable and natural audio output even when speech is present in the environment.
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 suppresses the influence of speech sounds, ensuring audio playback quality is maintained by accurately differentiating and removing speech components from noise, resulting in a more stable and natural sound experience.
Implementation Method 1
a microphone for receiving an external sound that includes the played-back audio sound and noise
Implementation Method 2
a noise-extracting unit that extracts a noise signal from an external sound signal detected by the microphone
Implementation Method 3
a speech-removing unit that removes a speech signal from the noise signal on the basis of noise spectrum data that is prepared in advance
Implementation Method 4
a correction unit that corrects an audio signal on the basis of a weighting factor
Implementation Method 5
an output unit that produces a played-back audio sound on the basis of the corrected audio signal
Data Source
AI summary
An audio apparatus has a function of correcting an audio signal in response to a noise level. The audio apparatus includes a correction unit that corrects an input audio signal on the basis of a weighting factor, an output unit that produces a played-back audio sound on the basis of the corrected audio signal, a microphone for receiving an external sound that includes the played-back audio sound and noise, a noise-extracting unit that extracts a noise signal from an external sound signal, the noise-extracting unit including a speech-removing unit that removes a speech signal from the noise signal on the basis of noise spectrum data, and a weighting factor calculation unit that calculates the weighting factor on the basis of the extracted noise signal and supplies the calculated weighting factor to the correction unit.