In-Device Privacy Framework for Smart Glasses
Solution Overview
Problem
Smart devices like smart glasses and smart watches pose privacy concerns as they can capture images and videos without users' awareness, leading to potential misuse of facial recognition technology, and existing solutions do not adequately protect individuals' privacy while allowing device functionality.
Innovation Solution
An in-device automated privacy framework that includes human detection, robust tracking, and intelligent de-identification of identity features, located near the operating system kernel to prevent manipulation, ensuring privacy while preserving image and video quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If face recognition and automated identification features are integrated into smart glasses, then the device can identify people and provide internet search results, but individuals' anonymity and privacy are compromised
Solution Approach 1:
The patent extracts and removes identifiable facial features from captured images through de-identification processing. The system separates identity-critical features from other image content, removing or obfuscating only the identifying portions while preserving the rest of the image quality and utility.
Solution Approach 2:
The patent applies de-identification processing before storing or sharing images. By proactively removing identifying features at the point of capture or immediately thereafter, the system prevents potential privacy breaches before they can occur, rather than attempting to address privacy concerns after images are distributed.
2Ease of operation
If smart glasses can capture photos and videos surreptitiously, then the wearer can record without being noticed, but it becomes impossible to recognize when recording is occurring
Solution Approach 1:
The patent introduces an intermediary privacy protection layer between the camera capture function and the stored/output images. This intermediary de-identification module acts as a mediator that processes captured images to remove identifying information, allowing the camera to function freely while the intermediary ensures privacy protection is applied automatically.
3Adaptability or versatility
If all captured data is stored in cloud servers, then data can be accessed and processed remotely, but privacy breaches and misuse of personal information increase
Solution Approach 1:
The patent extracts and removes personally identifiable information from images before they are stored in cloud servers. By taking out the identifying features at the point of capture and preventing them from being stored or transmitted to the cloud, the system enables cloud storage functionality while eliminating the primary source of privacy risk associated with storing personal images remotely.
4Productivity
If face recognition software is allowed to access captured images, then automated identification and information retrieval are enabled, but the risk of misuse and unauthorized access increases
Solution Approach 1:
The patent applies de-identification processing to images before they can be accessed by any software including face recognition systems. By preliminarily removing identifying features, the system prevents unauthorized access and misuse while still allowing automated processing of non-identifying image content for legitimate purposes.
Data Source
Figure 1~2
Figure 3a~3b(iv)
Figure 4
AI summary
The present invention relates to a method and a system for preserving the privacy of persons detected by a camera of a smart glass. The system comprises a camera module (301) for taking at least one image of a face by means of the camera of the smart glass and a detection module (302) for detecting a face in the image taken by the camera. A state module (303) is provided for generating a state signal if a face is detected (304) in the image. A tracking module (306) is provided for tracking features of the detected face in case a state signal is generated. Finally, the system comprises a de-identification module (307) for de-identifying identity features in the taken image, while keeping as much as possible information.