This invention describes a system for automatically adjusting the view during a video call. Specifically, it focuses on cropping the video to show only one person when they have a special status, like speaking or being highlighted, while excluding others. The system can then revert to showing multiple people once that special status changes, potentially triggered by a user's input.
Why it matters: Filed before the widespread adoption of highly efficient, on-device AI for real-time video analysis. Advances in computer vision and edge computing since 2021 make dynamic subject detection and intelligent cropping for video calls significantly more practical and less resource-intensive.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI