Video Feed Segmentation for Clear Participant Views in Conferences

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Video conferencing systems often present a single camera feed from a conference room, which can obscure participants and fail to clearly identify individual attendees, leading to lost visual information and confusion about who is present.

Innovation Solution

A system that segments the camera feed into multiple video feeds, which can be used in the field of computer systems, which can be used in the field of computer systems, which can be used in the field of video conferencing systems, specifically involving the segmentation of a camera feed into separate video streams for individual participants using facial recognition and metadata, allowing for clear identification and presentation of multiple attendees.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If a single camera feed is used to show all participants in a conference room, then the complete scene is captured, but individual participants cannot be clearly identified and visual information is lost

Engineering Contradiction:
Improvevisual informationVSAvoidvideo processing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent divides a single camera feed into multiple segmented video feeds, each showing a different participant. The system detects participants in the conference room and creates individual video streams by cropping and zooming in on each person, thereby preserving visual information while enabling clear identification of individual attendees without requiring multiple physical cameras.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If multiple cameras are deployed to capture individual participants, then clear identification of each attendee is achieved, but system complexity and cost increase

Engineering Contradiction:
Improveparticipant identification accuracyVSAvoidcamera system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

Instead of using multiple cameras, the patent uses a single camera feed that is computationally segmented into multiple individual participant views. The system processes the single video stream to detect and isolate each participant, achieving precise participant identification through software-based segmentation rather than hardware multiplication.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates multiple copied video feeds from a single source camera feed. Each participant receives a copied and customized version of the main video stream that is cropped and optimized to show only that participant, thereby achieving multiple views from a single physical camera through digital replication and processing.

Inventive Principle:
Principle #26Copying

3Loss of information

If individual video feeds are created for each participant, then attendee recognition is improved, but bandwidth and transmission requirements increase

Engineering Contradiction:
Improveattendee identification informationVSAvoidvideo data transmission
Core Design Contradiction:
Loss of informationVSQuantity of substance

Solution Approach 1:

The patent extracts individual participant information from the complete camera feed by cropping and isolating each person. Rather than transmitting entirely separate video feeds from multiple cameras, the system extracts and transmits only the relevant portions showing individual participants, reducing redundant data transmission while maintaining clear attendee identification.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS12519903B2Segmentation of video feed during video conference
Publication Date: 2026.01.06 LENOVO (SINGAPORE) PTE LTD
  • US12519903B2 patent drawing
  • US12519903B2 patent drawing
  • US12519903B2 patent drawing

AI summary

In one aspect, a first device includes a processor and storage. The storage includes instructions executable by the processor to facilitate a video conference and to segment a camera feed into first video and second video as part of facilitating the video conference. The first video shows a first person but not a second person, the second video shows the second person but not the first person, and the camera feed shows both the first and second people. The instructions are also executable to use the first video and the second video as part of the video conference. This might include transmitting the first and second video to a second device, and/or controlling the display of the second device to concurrently but separately present the first and second video.