Subject-Based Video Segmentation for Selective Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video communication sessions, participants often face the challenge of viewing unnecessary subjects within the entire field of view captured by smartphones, as traditional webcam solutions provide low-quality video and transmit the entire field of view, including sections that participants may not want to share.
Innovation Solution
A communication device equipped with an AI-driven subject-based video image segmentation module that delineates the video feed into primary and secondary segments, allowing selective transmission and viewing of specific subjects, enabling users to choose which segments to share with other participants based on accessibility settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the entire field of view is captured and transmitted, then all subjects are included in the video feed, but participants are exposed to unwanted content and privacy issues arise
Solution Approach 1:
The video feed is segmented into multiple regions of interest (ROIs), each containing a different subject. The system transmits only the selected ROI to remote participants, allowing the local participant to share only the desired portion of the video feed while maintaining simplicity in operation.
Solution Approach 2:
The system extracts and transmits only the selected region of interest from the full video feed, removing unwanted content before transmission. This extraction approach prevents exposure to unnecessary subjects while keeping the sharing process simple and automated.
2Manufacturing precision
If smartphone cameras are used to capture video, then video quality is improved, but the entire field of view includes sections participants may not want to share
Solution Approach 1:
The high-quality video feed from smartphone cameras is divided into multiple regions of interest. Each ROI contains a specific subject, and the system allows selection of which ROI to transmit, thereby maintaining video quality while preventing exposure to unwanted content.
Solution Approach 2:
Instead of transmitting the entire high-quality video feed, the system transmits only the locally selected region of interest with high quality, while other regions are excluded. This ensures that the transmitted content maintains smartphone camera quality without including unwanted sections.
3Adaptability or versatility
If the full video feed is transmitted to all participants, then everyone sees all subjects, but this reduces privacy control and customization options
Solution Approach 1:
The video feed is pre-segmented into multiple regions of interest before transmission. Each participant can then select which ROI to receive, providing privacy control and customization. The segmentation is performed once on the transmitting device, keeping processing complexity manageable.
Solution Approach 2:
The video feed is pre-processed into segmented regions of interest before transmission to participants. This preliminary segmentation allows participants to choose their preferred view in advance, enhancing privacy control without requiring complex real-time processing on receiving devices.
Data Source
AI summary
A communication device provides subject-based segmentation and selective presentation of a video feed. A processor identifies a primary subject and a secondary subject within a captured video stream. The processor delineates the video stream into a primary segment and at least one secondary segment respectively encompassing the primary and secondary subjects. The processor identifies for each connected second device, a request type from among: (i) a first request type to only receive the primary segment; (ii) a second request type to receive the primary segment and secondary segment(s); and (iii) a third request type to receive secondary segments, but not the primary segment. The processor transmits, to each second device, specific segments of the video feed based on the request type associated with each respective second device. An accessibility setting enables a second device to selectively receive video segments of a sign language interpreter along with a main presenter.