Participant-tuned filtering using deep neural network dynamic spectral masks for conversation isolation and safety in noisy environments

By generating and applying a composite spectrum mask, the problem that conversation participants have difficulty clearly hearing other participants in noisy environments is solved, achieving a clear and safe conversation isolation effect.

CN114746937BActive Publication Date: 2025-04-11INTERNATIONAL BUSINESS MACHINE CORPORATION
2 Cites 0 Cited by

Patent Information

Application Number
CN202080082205.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2019-12-02
Filing Date
2020-11-26
Publication Date
2025-04-11
Estimated Expiration
2040-11-26

AI Technical Summary

Technical Problem

In noisy environments, it is difficult for conversation participants to hear clearly and safely what other participants are saying, and prior art is difficult to effectively address this problem.

Method used

By receiving multiple spectrum masks, each corresponding to one participant in the conversation, a composite spectrum mask is generated and applied to the sound captured by the microphone to filter out mismatched sounds and amplify the matched sounds. This method combines deep neural networks to generate a personalized spectrum mask and share the spectrum mask between mobile devices to generate a composite spectrum mask.

Benefits of technology

The clarity and security of conversations between selected participants are achieved in noisy environments, improving conversation clarity and security.

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

Abstract

Isolating and amplifying a conversation between selected participants is provided. A plurality of spectral masks are received. Each spectral mask of the plurality of spectral masks corresponds to a respective participant of a selected group of participants included in the conversation. A composite spectral mask is generated by additive superposition of the plurality of spectral masks. The composite spectral mask is applied to sound captured by a microphone to filter out sounds that do not match the composite spectral mask and to amplify remaining sounds that match the composite spectral mask.
Need to check novelty before this filing date? Find Prior Art