Clicking Noise Detection in Digital Audio Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice recognition systems fail to reliably detect noise signals, particularly clicking noises, that occur in close proximity to speech signals due to their reliance on long quiet pauses, leading to incorrect detection or missed noise signals.
Innovation Solution
A method and device that divide digital audio signals into successive sections, evaluate energy content relative to a dynamically adjusted energy threshold, and count preceding and following sections to accurately detect high-energy signal patterns characteristic of noise signals, allowing for precise identification of clicking noises without requiring complex transformations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system relies on long quiet pauses to detect noise signals, then noise detection can be performed with simple methods, but noise signals occurring in close proximity to speech signals cannot be detected
Solution Approach 1:
The audio signal is divided into multiple signal sections (frames) of fixed duration, allowing the system to analyze energy content in discrete time segments. This segmentation enables detection of noise signals that occur within short time spans by examining energy patterns across multiple adjacent sections, rather than requiring long quiet pauses between speech signals.
Solution Approach 2:
The system dynamically evaluates energy contents of signal sections and compares them against energy thresholds to identify high-energy sections characteristic of noise. By counting the number of adjacent high-energy sections and evaluating their temporal pattern, the system can detect noise signals occurring in close proximity to speech, adapting to varying signal conditions without requiring fixed long pause periods.
2Measurement precision
If spectral analysis and CEPSTRAL representation are used for noise detection, then noise signals can be identified, but the computational complexity increases
Solution Approach 1:
The invention extracts only the essential energy content information from audio signal sections using simple energy calculation formulas, rather than performing full spectral analysis or CEPSTRAL transformation. This extraction of key energy characteristics maintains sufficient noise detection capability while significantly reducing computational complexity and processing requirements.
Data Source
AI summary
In a method (M) to detect a noise signal (PS1, PS2, PS3) in a digital audio signal (EAS), it is provided that the audio signal (EAS) is divided into successive signal sections (SAS), and the energy contents of successive signal sections (SAS) are determined, and the energy contents of a signal section (SAS) are evaluated in relation to an energy threshold (ET), and that the occurrence of at least one high-energy signal section having an energy content above the energy threshold (ET), and the occurrence of at least one signal section (SAS) preceding the at least one high-energy signal section and having an energy content below the energy threshold (ET), and the occurrence of at least one signal section (SAS) following the at least one high-energy signal section and having an energy content below the energy threshold (ET) are detected, and that a quantity of signal sections (SAS) that precede the at least one high-energy signal section and a quantity of high-energy signal sections and a quantity of signal sections (SAS) that follow the high-energy signal section are counted.