| CPC H04L 65/403 (2013.01) [G06N 20/00 (2019.01); G06V 30/153 (2022.01); G10L 21/10 (2013.01); G06V 30/148 (2022.01)] | 20 Claims |
|
1. A method comprising:
identifying, by a processor executing a machine-learning model, visual presentation content shared by a first electronic device of a first participant to a second electronic device of a second participant and a third electronic device of a third participant during an electronic communication session;
decoupling, by the processor, using an image segmentation protocol, a region of the visual presentation content from other visual elements shared during the electronic communication session;
executing an image recognition protocol on the region and not the other visual elements to identify a first set of words of the visual presentation content displayed within the region, wherein the first set of words is in a first language;
identifying, by the processor, a second preferred language of the second participant of the electronic communication session and a third preferred language of the third participant of the electronic communication session;
presenting, by the processor as a first overlay on at least a portion of the region during the electronic communication session on the second electronic device associated with the second participant, a second set of words in the second preferred language of the second participant, the second set of words corresponding to the first set of words;
presenting, by the processor as a second overlay on at least a portion of the region during the electronic communication session on the third electronic device associated with the third participant, a third set of words in the third preferred language of the third participant, the third set of words corresponding to the first set of words;
monitoring, by the processor, a set of pixels within the region to identify a visual revision within the region that satisfies a threshold of pixel differentiation between the set of pixels in a first pixel capture and the set of pixels in a second pixel capture; and
in response to identifying the visual revision:
executing, by the processor, the image recognition protocol to identify a fourth set of words within the region having the visual revision, wherein the fourth set of words is in the first language;
presenting, by the processor within the first overlay on the second electronic device, a fifth set of words, in the second preferred language, corresponding to the fourth set of words; and
presenting, by the processor within the second overlay on the third electronic device, a sixth set of words, in the third preferred language, corresponding to the fourth set of words.
|