Virtual Privacy Curtain With Depth-Based Unintended Participant Blurring

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Video conferencing systems often reveal the participant's location and expose unintended participants, compromising privacy and causing distractions in crowded environments.

Innovation Solution

Implementing a privacy curtain logic that uses machine learning to identify and blur or remove unintended participants from the video feed based on their distance and pose relative to the camera, creating a virtual privacy barrier.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If video conferencing is conducted in varied locations including crowded environments, then accessibility and convenience are improved, but privacy protection and reduction of distractions deteriorate

Engineering Contradiction:
Improveconvenience of video conferencingVSAvoidprivacy exposure and distractions
Core Design Contradiction:
Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The video feed is segmented into multiple regions based on depth information. The foreground region containing the intended participant is separated from the background region containing unintended participants. This segmentation allows the system to selectively process different regions with different opacity levels, maintaining the participant's visibility while obscuring the background to protect privacy and reduce distractions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different regions of the video feed are assigned different quality characteristics. The foreground region where the intended participant is located maintains full visibility and quality, while the background region applies varying degrees of obscuration. This local differentiation ensures that the participant's video quality is preserved while the harmful background elements are selectively treated.

Inventive Principle:
Principle #3Local quality

2Loss of information

If the camera captures the full scene including background participants, then completeness of visual information is improved, but privacy protection and focus on intended participant deteriorate

Engineering Contradiction:
Improvecompleteness of visual informationVSAvoidprivacy exposure
Core Design Contradiction:
Loss of informationVSObject-affected harmful factors

Solution Approach 1:

The system dynamically adjusts the opacity of background regions based on real-time depth information and identification of unintended participants. The obscuration level is not static but adapts to the current scene composition, maintaining information completeness for intended participants while dynamically protecting the privacy of unintended participants who appear in the background.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

A depth map serves as an intermediary data structure that mediates between the raw video feed and the final processed output. The depth map provides spatial information that enables the system to distinguish between foreground and background regions, allowing selective application of obscuration effects without losing the underlying visual information structure.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If machine learning models process video frames to identify participants, then accuracy of participant identification is improved, but computational complexity and processing time increase

Engineering Contradiction:
Improveaccuracy of participant identificationVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary actions by generating depth maps and identifying regions of interest before applying the full machine learning-based participant identification. This preliminary processing narrows down the search space and provides structural information that guides the more computationally intensive ML models, reducing their overall complexity and processing requirements while maintaining high identification accuracy.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS12356114B2Virtual privacy curtain
Publication Date: 2025.07.08 CISCO TECHNOLOGY INC
  • US12356114B2 patent drawing
  • US12356114B2 patent drawing
  • US12356114B2 patent drawing

AI summary

Presented herein are techniques to process an image for a video conference. A method includes obtaining an image from a camera, identifying a first person and a second person in the image, labeling pixels in the image belonging to the first person and to the second person, determining a first distance between the first person and the camera and a second distance between the second person and the camera, and based on the first distance and the second distance, selectively removing or blurring at least one of the first person and the second person from the image, before transmitting the image to an endpoint of a video conference, by manipulating the pixels in the image belonging to the at least one of the first person and the second person.