Eye Gaze Adjustment for Video Communication Misalignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video communications, presenters often read from a display device while their camera captures their face, leading to distracting eye movements and perceived gaze misalignment, which can reduce the effectiveness of communication.
Innovation Solution
A computing system that captures a video stream of a user, detects facial features, and computes a desired eye gaze direction to generate gaze-adjusted images, simulating natural eye movements like saccades, micro-saccades, and vergence, to maintain a natural and attentive appearance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the presenter reads from a display device during video communication, then the presenter can deliver structured information effectively, but the receiver perceives distracting eye movements and gaze misalignment
Solution Approach 1:
The patent creates synthetic eye images that copy and replace the presenter's actual eyes in the video feed. These synthetic eyes are generated to appear as if they are looking at the camera while the presenter reads from the display, effectively copying the desired gaze behavior without requiring the presenter to actually look at the camera.
Solution Approach 2:
The system introduces an intermediary processing layer between the camera capture and the video transmission. This intermediary component (the eye gaze adjustment module) analyzes the presenter's actual eye position and generates corrected eye images that mediate between the presenter's reading behavior and the receiver's perception, making the gaze appear aligned without disrupting the reading process.
2Device complexity
If the camera is positioned directly above the display device, then the setup is simplified and space is optimized, but the receiver perceives the presenter's eye gaze as being focused below eye level
Solution Approach 1:
The system generates synthetic eye images that copy the desired gaze direction (looking at the camera) and replaces the actual eye images. This allows the camera to remain in its simplified position above the display while the synthetic eyes create the perception of proper gaze alignment at eye level.
Solution Approach 2:
The system changes the parameter of eye gaze direction in the video feed by generating synthetic eyes with adjusted orientation. Instead of physically repositioning the camera or display, the solution modifies the gaze direction parameter through image synthesis, making the eyes appear to look at the camera even though the physical setup remains unchanged.
3Object-affected harmful factors
If the presenter maintains eye contact with the camera by looking directly at it, then the receiver perceives natural eye contact, but the presenter cannot read from the display device effectively
Solution Approach 1:
The system copies the presenter's actual eye appearance and behavior while replacing only the gaze direction. The synthetic eyes maintain the presenter's natural eye characteristics but are oriented to look at the camera, allowing the presenter to read from the display while the receiver sees natural-looking eye contact.
Solution Approach 2:
The eye gaze adjustment module acts as an intermediary that decouples the presenter's reading behavior from the receiver's perception of eye contact. The presenter can look at the display (ease of operation) while the intermediary generates synthetic eyes that appear to look at the camera (eye contact perception), resolving the conflict between these two requirements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computing system, a method, and a computer-readable storage medium for adjusting eye gaze are described. The method includes capturing a video stream including images of a user, detecting the user's face region within the images, and detecting the user's facial feature regions within the images based on the detected face region. The method includes determining whether the user is completely disengaged from the computing system and, if the user is not completely disengaged, detecting the user's eye region within the images based on the detected facial feature regions. The method also includes computing the user's desired eye gaze direction based on the detected eye region, generating gaze- adjusted images based on the desired eye gaze direction, wherein the gaze-adjusted images include a saccadic eye movement, a micro-saccadic eye movement, and/or a vergence eye movement, and replacing the images within the video stream with the gaze-adjusted images.