Removing visual content representing a reflection of a screen

A method, comprising:

• displaying, on a screen, a first screen content frame during a first time period;
• displaying, on the screen, a black frame during a second time period;
• wherein the black frame is a video frame that is black or empty;
• obtaining a target video captured by a sensor, wherein the target video includes a first video frame and a second video frame;
• wherein the first video frame is captured during the first time period during which the first screen content frame is displayed on the screen and the first video frame includes visual content representing a reflection of the first screen content;
• wherein the second video frame is captured during the second time period during which the black frame is displayed on the screen;
• generating a masking frame based on the first video frame and the second video frame;
• wherein generating the masking frame comprises:
• determining a clean video frame based on the second video frame;
• determining a tainted video frame based on the first video frame;
• generating the masking frame based at least on one or more deviations between the clean video frame and the tainted video frame;
• applying the masking frame to a target video frame, within the target video, to generate a modified target video frame;
• including the modified target video frame into a modified target video;
• wherein the method is performed by one or more devices, each including one or more hardware processors.

One or more non-transitory machine-readable media storing instructions which, when executed by one or more processors, cause:

• setting a screen to display a first screen content frame using a first polarization;
• displaying, on the screen, the first screen content frame using the first polarization during a first time period;
• setting a sensor to capture a first video frame using a second polarization, such that the sensor is able to capture a reflection of the first screen content frame displayed using the first polarization;
• capturing, by the sensor, the first video frame using the second polarization during the first time period;
• setting the screen to display a second screen content frame using a third polarization;
• displaying, on the screen, the second screen content frame using the third polarization during a second time period;
• setting the sensor to capture a second video frame using a fourth polarization, such that the sensor is unable to capture a reflection of the second screen content frame displayed using the third polarization;
• capturing, by the sensor, the second video frame using the fourth polarization during the second time period;
• responsive at least to determining that the first video frame is captured during the first time period during which the sensor is able to capture the reflection of the first screen content frame and that the second video frame is captured during the second time period during which the sensor is unable to capture the reflection of the first screen content frame:
• generating a masking frame based on the first video frame and the second video frame;
• wherein generating the masking frame comprises:
• determining a clean video frame based on the second video frame;
• determining a tainted video frame based on the first video frame;
• generating the masking frame based at least on one or more deviations between the clean video frame and the tainted video frame;
• applying the masking frame to a target video frame, within the target video, to generate a modified target video frame;
• including the modified target video frame into a modified target video.