Voice Recognition Compensation for IP Camera Noise Reduction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional internet Protocol cameras (IP cams) with noise reduction microphones often reduce voice recognition accuracy due to the lack of noise reduction in training datasets, necessitating a method to determine and compensate for noise in signals.

Innovation Solution

A method that transforms signals into spectrograms to identify sharp changes in frequency spectra, determines noise reduction by comparing the number of frames with sharp changes to a predetermined value, and compensates the signal by adding a noise signal to enhance voice recognition rates.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If noise reduction is performed in the microphone, then the signal quality is improved, but the voice recognition rate is reduced

Engineering Contradiction:

Improvesignal qualityVSAvoidvoice recognition rate

Core Design Contradiction:

Object-affected harmful factorsVSMeasurement precision

Solution Approach 1:

The system performs preliminary detection of noise reduction by analyzing sharp changes in frequency spectra before voice recognition. By transforming the signal to spectrogram and counting frames with sharp changes, the system identifies whether noise reduction has been applied, then prepares appropriate compensation in advance to maintain voice recognition accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system converts the harmful effect of noise reduction (loss of voice recognition accuracy) into a detectable pattern (sharp changes in frequency spectrum). By detecting these patterns and adding compensatory noise, the system transforms the originally harmful noise reduction effect into a manageable condition that can be corrected, ultimately benefiting voice recognition

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

2Object-affected harmful factors

If noise reduction is performed in the microphone, then the harmful factors are reduced, but loss of information occurs

Engineering Contradiction:

ImprovenoiseVSAvoidvoice signal information

Core Design Contradiction:

Object-affected harmful factorsVSLoss of information

Solution Approach 1:

The system implements feedback by detecting the presence of noise reduction through frequency spectrum analysis, then using this information to control the compensation process. The detected sharp changes in frequency spectra provide feedback about the signal processing history, enabling the system to adjust its compensation strategy accordingly to recover lost voice information

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11380346B2Signal processing system and a method of determining noise reduction and compensation thereof

Publication Date: 2022.07.05 WISTRON CORP

AI summary

A method of determining noise reduction in a signal includes transforming the signal to generate a spectrogram; determining sharp change in a frequency spectrum for each frame in the spectrogram; and comparing a counted number of frames having sharp change with a predetermined value. The signal is determined to be subject to noise reduction if the counted number is greater than the predetermined value.