Wearable Face Capture Triggering for Accurate 3D Avatar Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wearable devices struggle to accurately capture images of a user's face for generating a realistic virtual avatar due to limitations in field of view and positioning of inward- and outward-facing imaging systems, leading to difficulties in rendering natural facial expressions and movements.
Innovation Solution
The wearable device employs an inward-facing imaging system to capture face images when the device is being put on or taken off, using eye cameras to synthesize a 3D face model without requiring additional user actions, and automatically starts and stops imaging based on device movement triggers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the wearable device uses inward-facing imaging system to capture face images, then the device can capture facial features, but the field of view is limited and cannot capture the entire face accurately
Solution Approach 1:
The system captures facial images during the process of putting on or taking off the wearable device, when the device is not yet positioned on the user's head. This preliminary action allows the imaging system to capture the entire face with adequate field of view before the device's positioning constraints are applied.
Solution Approach 2:
The imaging system dynamically adjusts its operation based on device movement detection. The controller activates the imaging system only when movement is detected (during donning/doffing), and deactivates it when the device is stationary on the user's head, optimizing both image capture quality and energy consumption.
2Measurement precision
If the wearable device requires additional user actions to capture face images, then the imaging can be more controlled, but the ease of operation decreases
Solution Approach 1:
The wearable device automatically detects when it is being put on or taken off using its movement sensors, and autonomously activates the imaging system to capture facial images during this process. The user simply needs to perform the natural action of putting on or removing the device, without any additional commands or actions required.
Solution Approach 2:
The system uses feedback from movement sensors to automatically control the imaging system. When movement exceeding a threshold is detected, the controller activates image capture; when movement ceases, the controller stops capture. This closed-loop control eliminates the need for user intervention while ensuring images are captured at the optimal moment.
3Measurement precision
If the wearable device continuously captures face images, then the facial data can be more complete, but the energy consumption increases
Solution Approach 1:
Instead of continuous imaging, the system uses periodic action triggered by movement detection. The imaging system is activated only during specific periods when the device is being put on or taken off, and remains inactive during normal wear, dramatically reducing energy consumption while capturing sufficient facial data.
Solution Approach 2:
The system performs the energy-intensive imaging operation in advance, during the donning/doffing process, before the device is worn. This preliminary capture of facial images eliminates the need for continuous imaging during extended wear periods, optimizing the energy efficiency of the device.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods (1200) for generating a face model for a user of a head-mounted device are disclosed, comprising detecting (1210), by the IMU, a trigger for imaging a face of the user, wherein the trigger comprises determining that an acceleration of the display device passes a threshold acceleration; capturing (1220), by the imaging system, images of at least a portion of a face of the user; analyzing (1230) the images captured by the imaging system; and generating (1240) the face model based at least partly on the analysis of the images..