Video Conference Face Assignment That Ignores Displayed Remote Faces

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video conferencing systems erroneously assign faces displayed on a video display screen to separate windows, despite those faces not being physically present in the conference room, leading to user confusion and distraction.

Innovation Solution

Implement a system that identifies displays visible to the video camera and determines if a face on the display corresponds to a non-present participant, preventing the assignment of such faces to separate windows by using facial analysis and machine learning models to distinguish between local and remote participants.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If face detection techniques are used to automatically extract and isolate participant faces to separate windows, then the ability to focus attention on particular persons is improved, but faces displayed on video screens are erroneously assigned to separate windows causing user confusion

Engineering Contradiction:

Improveability to focus attention on participantsVSAvoidaccuracy of face assignment

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The patent introduces an intermediary verification process that checks whether detected faces are actually present in the conference room before assigning them to separate windows. This intermediary step prevents the erroneous assignment of faces that are merely displayed on video screens, thereby resolving the contradiction between ease of operation and reliability.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback mechanisms that continuously monitor and verify the physical presence of participants. By using multiple data sources (video feeds, presence detection) and comparing them against face detection results, the system provides feedback to correct potential errors in face assignment, thus improving reliability while maintaining ease of operation.

Inventive Principle:

Principle #23Feedback

2Productivity

If auto-framing operations extract faces from video feeds, then individual participant windows are created for focused viewing, but faces on display screens are mistakenly identified as local participants

Engineering Contradiction:

Improveefficiency of participant presentationVSAvoidmisidentification of participant status

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent segments the face detection and assignment process into distinct stages: initial face detection, verification of physical presence, and final window assignment. This segmentation allows the system to handle the complex task of distinguishing between local and remote faces systematically, improving productivity while preventing misidentification.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary verification of participant presence before conducting face extraction and window assignment. By checking whether faces are actually present in the conference room beforehand, the system prevents the loss of information regarding participant status, ensuring that only genuine local participants are assigned separate windows.

Inventive Principle:

Principle #10Preliminary action

3Area of stationary object

If the system assigns separate windows to all detected faces, then every face receives dedicated display space, but non-present participants are incorrectly included causing distraction

Engineering Contradiction:

Improvedisplay space for participantsVSAvoiduser confusion and distraction

Core Design Contradiction:

Area of stationary objectVSObject-affected harmful factors

Solution Approach 1:

The patent applies local quality by treating faces differently based on their verification status. Faces confirmed to be physically present in the conference room are assigned separate windows with dedicated display space, while faces merely displayed on screens are excluded. This differentiated approach ensures that display space is allocated appropriately without causing user confusion.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS12489868B2Preventing faces detected on a video display screen from assignment to separate window during video conference

Publication Date: 2025.12.02 CISCO TECHNOLOGY INC

AI summary

Video is obtained in a video conference session that includes one or more participants in a video conference room, wherein the video conference room includes a video display and a one video camera, and wherein a participant is remote with respect to the video conference room. During the video conference session, it is determined that video captured by the camera in the video conference room includes a face of a person not physically present in the video conference room and which face is displayed on the video display. Video of each participant in the video conference room is assigned to a respective video layout window except for video of the face of the person not physically present in the video conference room that is displayed on the video display.