Selective image broadcasting in a video conference
Video conferencing systems are popular means for remote parties to interact in real-time. A user may wish to emphasize certain visual elements of the video conference, for example, a company's logo or the contents of a whiteboard may be important to remain clearly seen by the viewers of the video conference. As elements are obscured, a previously captured image of the obscured visual element may be overlaid on top of the obscuring image to visual remove the obscuring object. Additionally, certain visual elements may be enhanced, such that a small visual element may appear larger or clearer within the video image.
1. A system for presenting a modified video image, comprising:
a camera;
a processor;
a network interface to a network;
a data storage; and
wherein the processor:
receives a video image comprising a number of visual elements captured within a field of view of the camera wherein, at a first time, the video image captures a first element of the number of visual elements;
accesses a record in the data storage comprising indicia of important visual elements;
upon determining, at a second time after the first time, that the first element of the number of visual elements matches the record, accessing an enhancement operation associated with the first element;
enhance the video image to produce an enhanced video image comprising the first element, as enhanced by the enhancement operation, and wherein the enhancement is not applied to an entirety of the video image and wherein the enhancement operation overlays the previously captured video image of the first element on top of an obscuring visual feature that obscures the first element; and
transmit, via the network interface, the enhanced video image to at least one recipient device.
2. The system of claim 1 , wherein the enhancement operation comprises the processor de-enhancing the video image to reduce bandwidth utilized to transmit the enhanced video image, wherein the first element is excluded from the de-enhancing operation.
3. The system of claim 1 , wherein the processor overlays the previously captured video image with a degree of transparency to allow the obscuring visual feature to remain partially visible.
4. The system of claim 1 , wherein the enhancement operation comprises the processor overlaying a previously captured image of the first element, on top of a currently captured image of the first element, wherein the previously captured image has been visually enhanced.
5. The system of claim 4 , wherein the previously captured image has been visually enhanced comprises one or more contrast modification, saturation modification, hue modification, brightness modification, color alteration, rotation, de-rotation, or sharpening.
6. The system of claim 1 , wherein:
the record comprises a boundary-limited portion of the first element previously captured and a location of the first element relative to a second element of the number of visual elements; and
the processor, upon determining the video image comprises at least a portion of the second element and the location of the first element relative to the second element, overlays the first element at the location.
7. The system of claim 6 , wherein:
the second element comprises a number of views corresponding to variations of at least one of rotation or translation of the second element within the video image;
the location of the first element is relative to ones of the number of views of the second element; and
the processor determines the video image comprises at least the portion of the second image and the location, further comprises the processor determining the video image comprises a current view matching one of the number of views of the second element and the location relative thereto and, in response to the determination, overlays the first element at the location relative to the current view of the second element.
8. The system of claim 7 , further comprising the processor altering the first element similarly to the at least one of rotation or translation of the second element within the video image.
9. The system of claim 1 , wherein the processor determines the first element matches the record comprising, matching the first element to a neural network trained to recognize the first element.
10. The system of claim 9 , wherein the neural network is trained, the training comprising:
accessing a set of data entries that comprise the first element;
applying one or more transformations to the set of data entries including rotation, translation, color alteration, contrast alteration, sharpness alteration to create a set of modified data entries;
creating a first training set comprising a collected set of data entries, the modified set of data entries, and a set of non-first element data entries;
training the neural network in a first stage of training using the first training set;
creating a second training set for a second stage of training comprising the first training set and the set of non-first element data entries that are incorrectly detected as data entries that comprise the first element in the first stage of training; and
training the neural network in a second stage using the second training set.
11. A method, comprising:
receiving a video image comprising a number of visual elements captured within a field of view of a camera wherein, at a first time, the video image captures a first element of the number of visual elements;
accessing a record in a data storage comprising indicia of important visual elements;
upon determining, at a second time after the first time, that the first element of the number of visual elements matches the record, accessing an enhancement operation associated with the first element;
enhancing the video image to produce an enhanced video image comprising the first element, as enhanced by the enhancement operation, and wherein the enhancement operation is not applied to an entirety of the video image; and
transmitting, via a network interface to a network, the enhanced video image to at least one recipient device.
12. The method of claim 11 , wherein the enhancement operation comprises de-enhancing the video image to reduce bandwidth utilized to transmit the enhanced video image, wherein the first element is excluded from the de-enhancing operation.
13. The method of claim 11 , wherein overlaying the previously captured video image comprises overlaying the previously captured video image with a degree of transparency to allow an obscuring visual feature to remain partially visible.
14. The method of claim 11 , wherein enhancing comprises overlaying a previously captured image of the first element, on top of a currently captured image of the first element, wherein the previously captured image has been visually enhanced.
15. The method of claim 14 , wherein the previously captured image has been visually enhanced comprising one or more of contrast modification, saturation modification, hue modification, brightness modification, color alteration, rotation, de-rotation, sharpening, or object substitution.
16. The method of claim 11 , further comprising:
the record, further comprising a boundary-limited portion of the first element previously captured and a location of the first element relative to a second element of the number of visual elements; and
upon determining the video image comprises at least a portion of the second element and the location of the first element relative to the second element, overlaying the first element at the location.
17. The method of claim 16 , further comprising:
the second element comprising a number of views corresponding to variations of at least one of rotation or translation of the second element within the video image;
the location of the first element is further relative to ones of the number of views of the second element;
altering the first element similarly to the at least one of rotation or translation of the second element within the video image and
upon determining the video image comprises at least the portion of the second image and the location, further comprises, upon determining the video image comprises a current view matching one of the number of views of the second element and the location relative thereto and, in response to the determination, overlaying the first element at the location relative to the current view of the second element.
18. A video conferencing system, comprising:
a communication interface to a camera;
a processor;
a network interface to a network;
a data storage; and
wherein the processor:
receives, from the camera, a video image comprising a number of visual elements captured within a field of view of the camera wherein, at a first time, the video image captures a first element of the number of visual elements;
accesses a record in the data storage comprising indicia of important visual elements;
upon determining, at a second time after the first time, that the first element of the number of visual elements matches the record, accesses an enhancement operation associated with the first element;
enhances the video image to produce an enhanced video image comprising the first element, as enhanced by the enhancement operation, and wherein the enhancement operation is not applied to an entirety of the video image; and
transmits via the network interface, conference content to at least one recipient device, wherein the conference content comprises the enhanced video image.
19. The system of claim 18 , wherein:
the record, further comprises a boundary-limited portion of the first element previously captured and a location of the first element relative to a second element of the number of visual elements; and
the processor, upon determining the video image comprises at least a portion of the second element and the location of the first element relative to the second element, overlays the first element at the location.
20. The system of claim 19 , further comprising:
the second element comprising a number of views corresponding to variations of at least one of rotation or translation of the second element within the video image;
the location of the first element is further relative to ones of the number of views of the second element;
the processor alters the first element similarly to the at least one of rotation or translation of the second element within the video image; and
upon determining the video image comprises at least the portion of the second image and the location, further comprises, upon determining the video image comprises a current view matching one of the number of views of the second element and the location relative thereto and, in response to the determination, overlaying the first element at the location relative to the current view of the second element.