Gaze Correction System for Video Conferencing Eye Contact

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In video conferences, the lack of eye contact due to the speaker's gaze not being directed at the camera creates a perceived disconnect, affecting non-verbal communication and making virtual meetings feel less immersive compared to face-to-face interactions.

Innovation Solution

The system corrects the recording perspective to make it appear as though the speaker is looking directly at the camera, using multiple cameras to determine and adjust the viewing direction, ensuring eye contact is maintained, and applying automated perspective corrections to align the speaker's gaze with the center of the screen.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If the recording device is positioned outside the display device, then the speaker can see other participants on the screen, but the speaker's gaze does not align with the camera, creating a lack of perceived eye contact

Engineering Contradiction:
Improvespeaker's ability to see other participantsVSAvoidalignment between speaker's gaze and camera recording
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent introduces correction means as an intermediary component that processes the recording from the camera positioned outside the display. This correction means digitally adjusts the recording to simulate eye contact, effectively mediating between the physical camera position and the desired visual effect of direct eye contact with other participants.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent applies parameter changes by digitally modifying the recording parameters (position, orientation, perspective) to create the effect of eye contact. The correction means alters the spatial parameters of the recorded image to compensate for the physical separation between camera and display positions.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If multiple recording devices are used to capture the speaker from different angles, then better perspective correction can be achieved, but the device complexity increases

Engineering Contradiction:
Improveperspective correction accuracyVSAvoidnumber of recording devices
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent segments the recording function by using multiple recording devices positioned at different locations (e.g., integrated camera, external camera, webcam). Each device captures the speaker from its specific angle, and the system selects or combines these segmented recordings to achieve optimal perspective correction.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent makes the recording system universal by designing it to accept and process recordings from multiple different device types and positions. The correction means is configured to handle various input sources and automatically determine the best perspective correction based on the available recordings.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4329291A1System, method, computer program and computer-readable medium
Publication Date: 2024.02.28 TELEFONICA GERMANY GMBH & CO
  • EP4329291A1 patent drawingFigure 1
  • EP4329291A1 patent drawingFigure 2
  • EP4329291A1 patent drawingFigure 3

AI summary

The present invention relates to a system for capturing and displaying at least a part of a person's face, wherein the system comprises a recording means, preferably with an optical center, correction means and a display means, wherein the recording means is configured to take a picture of at least a part of the person's face, wherein the correction means are configured to modify the picture and wherein the display means is configured to display at least a part of the modified picture, wherein the correction means are configured to modify the picture in such a way that the gaze direction of the person shown on the display means is shown such that the person is looking into the recording means, preferably at the optical center of the recording means.