Audio analysis method and system based on decibel and energy value conversion

By combining the analysis of decibel values ​​and energy values, and using the deep neural network model, the accuracy and resource consumption problems of existing audio analysis technologies in complex environments and nonlinear signal processing are solved, and more efficient audio signal analysis is achieved.

CN119028371AInactive Publication Date: 2024-11-26NANJING LONGYUAN INFORMATION TECH CO LTD
View PDF 12 Cites 0 Cited by

Patent Information

Application Number
CN202410969187.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing audio analysis techniques have challenges in analytical accuracy and resource consumption when dealing with complex audio environments and nonlinear, non-stationary signals, and ignore the correlation between decibel values ​​and energy values.

Method used

Using an audio analysis system based on decibel and energy value conversion, the audio signal preprocessing, decibel value calculation and energy value extraction module is combined with the deep neural network model of attention mechanism, and the decibel value and energy value are fused to achieve the update of the audio signal. Comprehensive and in-depth analysis.

Benefits of technology

Improves the accuracy and reliability of audio analysis, reduces resource consumption, and shows excellent performance in real-time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119028371A_ABST
    Figure CN119028371A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of audio analysis, in particular to an audio analysis method and system based on decibel and energy value conversion. The audio signal preprocessing module is used for receiving an audio signal, preprocessing the audio signal, and removing noise in the audio signal to obtain an audio processing signal; the decibel value calculation module is used for calculating an SPL value in the audio processing signal and extracting a decibel value; the energy value extraction module is used for calculating a short-time energy value of each frame of audio processing signal; the conversion fusion module is used for fusing the decibel value and the energy value, and integrating and extracting key information in the audio signal; the analysis of the decibel value and the energy value is combined to realize more comprehensive and deep analysis of the audio signal, so that the accuracy and the reliability of audio analysis can be improved, and the resource consumption can be reduced while the real-time performance is ensured.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Substation equipment sound and voice identification method

    CN104409073A

  • Lie detection method based on fixed duration voice emotion recognition sequence analysis

    CN110265063A

  • Figure image generating method, figure image interaction method, figure image generating device, figure image interaction device and terminal equipment

    CN110648672A

  • Transformer substation online intelligent inspection system and method

    CN112350441A

  • MFCC coefficient-based pig sound recognition system and method

    CN114203187A