US 12,323,263 B2
Gaze repositioning during a video conference
Razvan Condorovici, Bucharest (RO); and Andra Stan, Bucharest (RO)
Assigned to Tobii AB, Danderyd (SE)
Filed by Tobii Technologies Limited, Galway (IE)
Filed on Jun. 14, 2023, as Appl. No. 18/209,893.
Application 18/209,893 is a continuation of application No. 17/495,800, filed on Oct. 6, 2021, granted, now 11,722,329.
Prior Publication US 2023/0327897 A1, Oct. 12, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. H04L 12/18 (2006.01); G06V 20/40 (2022.01); G06V 40/18 (2022.01)
CPC H04L 12/1831 (2013.01) [G06V 20/46 (2022.01); G06V 40/18 (2022.01); H04L 12/1822 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A method at a first client conferencing system associated with a first participant of a videoconference and operably in communication with at least a second client conferencing system and a third client conferencing system associated with a second participant and a third participant, respectively, of the videoconference, the method comprising:
generating first metadata that includes that information that associates the first participant, the second participant, and the third participant with an adjusted eye gaze for the first participant based on a location of the second participant and the third participant on a first display on the first client conferencing system; and
modifying an eye region of the first participant according to the generated first metadata on a second display on the second client conferencing system and on a third display on the third client conferencing system;
wherein the first client conferencing system comprises at least a first display, and the second client conferencing system comprises a first video camera and a second display, and wherein the third client conferencing system comprises a second video camera, and wherein the method further comprising receiving, from the third client conferencing system, a second video signal of the third participant, a second video signal being acquired by the second video camera; and displaying the received second video signal on a second area of the first display;
wherein the first client conferencing system further comprises a third video camera, and wherein the method further comprises acquiring, by the third video camera, a third video signal, the third video signal comprising at least one second video frame including an image of the first participant looking at a fourth area of the first display configured to display a video signal of a fourth participant of the videoconference;
determining, from the image, a gaze direction of the first participant toward a position within the fourth area of the first display;
generating second metadata associated with said at least one second video frame and including, based on the determined gaze direction, an identity of the fourth participant of the videoconference; and
sending the second metadata with the associated second video frame to at least one client conferencing system associated with a participant of the videoconference;
wherein: either the fourth participant corresponds to said second participant, and the fourth area of the first display corresponds to a first area configured to display the first video signal; or the fourth participant corresponds to said third participant and the fourth area of the first display corresponds to said second area configured to display the second video signal; and
wherein sending the second metadata with the associated second video frame to at least one client conferencing system associated with a participant of the videoconference comprises sending the second metadata to at least the third client conferencing system, if the second metadata includes an identity of the second participant, or sending the second metadata to at least the second client conferencing system, if the second metadata includes an identity of the third participant; and
wherein determining the gaze direction of the first participant toward a position within the fourth area of the first display comprises: using stored gaze directions of the first participant toward the corners of the first display, the stored gaze directions being determined from images of the first participant acquired by the third video camera during a calibration phase.