US Patent 12069109 Virtual background rendering based on target participant view

A local background replacement image (LBGRI) is received at a first device of a first conference participant. The first device is connected to an audio-visual conference that includes conference participants. The conference participants include the first conference participant and a second participant. A source output image that includes a source foreground segment representative of the second participant is received from a second device of the second participant. The LBGRI is combined with the source foreground segment to generate a target output image. The target output image is output at the first device.