Screen Share Obscuration via ML Content Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing screen sharing technologies do not effectively obscure unwanted content during screen sharing sessions, potentially revealing private or embarrassing information to others.
Innovation Solution
A method and apparatus that utilize machine learning algorithms to identify and obscure specific portions of a user's view, allowing users to control and train the obscuration of display elements, ensuring only intended content is shared during real-time screen sharing sessions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the entire desktop view is shared during a screen sharing session, then all content is visible to other users, but private or unwanted content is exposed
Solution Approach 1:
The system performs preliminary identification of content types and user preferences before the screen sharing session begins. User preferences for obscuring specific content types (e.g., notifications, ads, private windows) are pre-configured, and the ML model is pre-trained to recognize these content types, enabling automatic obscuration without real-time manual intervention
Solution Approach 2:
A machine learning model acts as an intermediary between the screen capture system and the display system. This ML intermediary automatically analyzes the captured screen content, identifies unwanted elements based on pre-configured preferences, and determines which portions should be obscured, eliminating the need for complex real-time user controls
2Ease of operation
If manual controls are provided for obscuring content, then user control is enhanced, but ease of operation decreases
Solution Approach 1:
The system provides self-service by automatically identifying and obscuring unwanted content without requiring user intervention during the screen sharing session. The ML model continuously monitors the screen content and autonomously applies obscuration based on pre-configured user preferences, making the system easy to operate while maintaining privacy
Solution Approach 2:
The system changes the parameter of obscuration from a static manual configuration to a dynamic automated process. User preferences are stored as configurable parameters that define what content types should be obscured, and the ML model adjusts the obscuration in real-time based on these parameters without requiring complex user controls
3Extent of automation
If machine learning algorithms are used to identify content, then automation is increased, but device complexity increases
Solution Approach 1:
The machine learning model is pre-trained offline to recognize various content types (notifications, advertisements, private windows, etc.) and user preferences for obscuring them. This preliminary training eliminates the need for complex real-time processing and manual configuration, achieving high automation with manageable system complexity
Solution Approach 2:
The system uses a trained ML model that has learned from training data to identify content types, effectively copying the expertise of manually trained operators into an automated system. The model captures patterns from training examples and applies them automatically during screen sharing, achieving high automation without proportionally increasing system complexity
Data Source
AI summary
The technology disclosed herein enables automatic obscuration of a portion of a view shared during a screen sharing session. In a particular embodiment, a method includes receiving a view displayed by a first user system to a first user. The method further includes identifying a portion of the view not to be shared and obscuring the portion of the view to generate a shared view. Also, the method includes transmitting the shared view to a second user system over a real-time screen sharing session, wherein the second user system displays the shared view to a second user.