3D Perspective Control for Video Conferencing Eye Contact
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video conferencing experiences are often unengaging due to difficulties in maintaining eye contact and unflattering views, primarily caused by the offset positioning of inexpensive webcams which capture images from angles, resulting in a 'fisheye effect' and lack of natural appearance.
Innovation Solution
Employing virtual cameras positioned in 3D space based on the location of participants on the display screen, using depth information from physical cameras to render a more flattering and head-on view, simulating a telephoto effect by positioning the virtual camera behind the display screen.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the video camera is positioned offset from the display screen to avoid blocking the view, then the display screen is accessible, but the camera captures images from an angle causing loss of eye contact
Solution Approach 1:
The patent creates a virtual camera that copies the function of a physical camera but positions it virtually behind the display screen. This virtual copy captures images from the correct head-on perspective without physically blocking the screen, resolving the contradiction between screen accessibility and eye contact maintenance.
Solution Approach 2:
The patent introduces a virtual camera as an intermediary between the physical camera and the final image output. The virtual camera receives images from the physical offset camera and computationally repositions them to appear as if captured from behind the screen, mediating between the physical constraint and the desired visual effect.
2Measurement precision
If the video camera is positioned close to the user to capture detailed images, then image detail is improved, but the wide-angle lens creates a fisheye effect causing unflattering appearance
Solution Approach 1:
The patent changes the effective focal length parameter by using a virtual camera positioned farther away from the user. This parameter change transforms the wide-angle fisheye perspective into a more flattering telephoto-like view while preserving the ability to capture detailed images through computational processing.
3Ease of operation
If the video camera is positioned behind the display screen to achieve head-on views, then eye contact is maintained, but the physical screen blocks the camera view
Solution Approach 1:
The patent creates a virtual copy of the camera positioned behind the screen, eliminating the need for complex physical modifications like transparent screens or multiple cameras. The virtual camera achieves the desired positioning through software without physical constraints.
Solution Approach 2:
The patent replaces the mechanical solution of physically positioning a camera behind the screen with a computational approach. Instead of modifying the physical setup, the system uses image processing and virtual camera techniques to achieve the same visual effect.
4Ease of operation
If a transparent screen is used to mount the camera behind it, then head-on views are achieved, but the solution does not support multi-user configurations
Solution Approach 1:
The patent creates a universal solution that works for both single-user and multi-user configurations. The virtual camera system can dynamically adjust to different user arrangements and screen layouts, providing head-on views regardless of the number of participants or their positions on the screen.
Data Source
AI summary
In one embodiment, images of a first user in a video conference are captured with one or more physical video cameras. The captured images are processed to form a three-dimensional (3-D) model of the first user. A location on a display screen is determined where an image of each of one or more second users in the video conference is shown. One or more virtual cameras are positioned in 3-D space. Each virtual camera is associated with a respective second user and positioned in 3-D space based on the location on the display screen where the image of the associated second user is shown. A view of the first user from the perspective of each of the one or more virtual cameras is rendered. The rendered view of the first user from the perspective of each virtual camera is shared with the associated second user for the respective virtual camera.