Selective AI Audio Processing for Virtual Meeting Understandability

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional virtual meeting platforms lack the capability to modify audio data for participants with speech or hearing issues, leading to reduced understandability and a degraded user experience.

Innovation Solution

Implementing artificial intelligence models to modify audio data in virtual meetings, enhancing understandability by generating modified audio streams to address speech and hearing challenges.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If audio data is transmitted without modification in virtual meetings, then the system maintains simplicity and low processing overhead, but participants with speech or hearing issues experience reduced understandability and discomfort

Engineering Contradiction:
ImproveunderstandabilityVSAvoidaudio processing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

An AI model is introduced as an intermediary component between the audio stream source and the participants. The AI model receives the original audio stream, processes it to enhance speech clarity and correct speech issues, and outputs a modified audio stream. This intermediary layer adds the necessary processing complexity only where needed to improve understandability for participants with speech or hearing impairments, without requiring changes to the entire virtual meeting system architecture.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If AI models are implemented to modify audio streams in real-time, then understandability for participants with impairments is improved, but computational resources and processing time are increased

Engineering Contradiction:
ImproveunderstandabilityVSAvoidcomputational resource consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The audio processing is applied selectively rather than uniformly to all audio streams. The system identifies specific audio streams that require modification (those from participants with speech issues or intended for participants with hearing impairments) and applies AI processing only to those specific streams. This local application of processing reduces overall computational resource consumption compared to processing all audio streams universally, while still achieving the understandability improvement where needed.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4632736A1Modifying audio data in a virtual meeting to increase understandability
Publication Date: 2025.10.15 GOOGLE LLC
  • EP4632736A1 patent drawingFigure 1
  • EP4632736A1 patent drawingFigure 2
  • EP4632736A1 patent drawingFigure 3

AI summary

A method for modifying audio data in a virtual meeting to increase understandability includes causing a virtual meeting UI to be presented during a virtual meeting between one or more participants. The virtual meeting UI provides first audio data associated with an audio stream produced by a client device of a first participant of the one or more participants. The method includes determining that the first audio data is to be modified during the virtual meeting. The method includes generating, using an AI model and using the audio stream produced by the client device of the first participant as input to the AI model, a modified audio stream to improve understandability of the first audio data by one or more participants. The method includes causing second audio data associated with the modified audio stream to be provided during the virtual meeting in place of the first audio data.