Adaptive noise reduction method and system for telephone traffic earphone based on environmental sound detection

By using multi-dimensional environmental noise analysis and online reinforcement learning to optimize state decision mapping parameters, the problem of noise suppression and voice fidelity in call center headsets under complex acoustic scenarios was solved, achieving improved adaptive noise reduction and voice quality.

CN122269186APending Publication Date: 2026-06-23BEIJING ZHONGLIAN NORTH INFORMATION TECH CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202610639584.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-05-11
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Existing noise reduction technologies in call center headsets are ill-suited to adapting to dynamically changing environmental noise, resulting in voice distortion and processing delays. They fail to achieve the optimal balance between noise suppression and voice fidelity in complex acoustic scenarios.

Method used

The system acquires the raw mixed signal through the built-in microphone of the headset, performs multi-dimensional environmental noise analysis, generates critical control commands using a state decision mapper, and combines online estimation and gain stabilization control to achieve perceptual domain spectrum enhancement processing and online reinforcement learning, thereby optimizing the state decision mapping parameters.

Benefits of technology

It improves the adaptive noise reduction capability of the headset in complex environments, enhances voice quality and user auditory comfort, and resolves the conflict between environmental noise suppression and call voice fidelity in real-time signal processing.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a traffic earphone adaptive noise reduction method and system based on environmental sound detection, and belongs to the technical field of communication acoustic signal processing. Making a decision through multi-dimensional environment noise analysis and based on the state decision mapping parameters of the state decision mapper to obtain a critical control instruction; obtaining an intermediate audio signal based on the original mixed signal through critical stable point online estimation and gain stability control; enhanced speech signals are obtained through perception domain spectrum enhancement processing; according to the method, on the basis of noise analysis, the instruction is generated through the mapper, stability control and spectrum enhancement are executed, parameters are optimized through reinforcement learning, active stability control over an acoustoelectric feedback closed loop is achieved, and the stability of the acoustoelectric feedback closed loop is improved. The conflict between noise suppression and voice fidelity is solved, and the adaptive noise reduction capability and the voice quality are improved.
Need to check novelty before this filing date? Find Prior Art