US 12,445,574 B2
Identification of persons of interest in meetings
Pei Hsuan Li, Spring, TX (US); Rose Hedderman, Spring, TX (US); Yu Chun Huang, Spring, TX (US); and Sarah Shiraz, Spring, TX (US)
Assigned to Hewlett-Packard Development Company, L.P., Spring, TX (US)
Filed by HEWLETT-PACKARD DEVELOPMENT COMPANY, L.P., Spring, TX (US)
Filed on Oct. 27, 2022, as Appl. No. 17/975,417.
Prior Publication US 2024/0146874 A1, May 2, 2024
Int. Cl. H04N 7/15 (2006.01); G06V 10/70 (2022.01); G06V 40/16 (2022.01); G10L 17/02 (2013.01); H04L 12/18 (2006.01)
CPC H04N 7/152 (2013.01) [G06V 10/70 (2022.01); G06V 40/171 (2022.01); G06V 40/172 (2022.01); G10L 17/02 (2013.01); H04L 12/1822 (2013.01)] 22 Claims
OG exemplary drawing
 
1. An electronic device, comprising:
a processor to:
receive multimodal content associated with a meeting space, the multimodal content comprising image data, audio data or video data;
identify, in the multimodal content, a behavioral trait of a person present in the meeting space;
apply, to the multimodal content, a machine learning model trained on behavioral profiles of individuals labeled as leaders;
perform, by the machine learning model, a comparison between the behavioral trait and the behavioral profiles;
determine, based on the comparison, whether the person is a leader of a meeting; and
based on a comparison of the facial feature, a gesture, and a speech pattern of the person with facial features, gestures, and speech patterns of leaders,
modify, based on the determination, a manner in which of the person is displayed presented on a graphical user interface (GUI).
 
10. An electronic device, comprising:
a processor to:
receive multimodal content comprising an image and audio data of a meeting space;
compare a facial representation of a person in the meeting space a stored entry in a database to determine whether the person is a leader te-of a meeting based on a match between the facial representation and the stored entry;
wherein responsive to the a comparison of the facial representation indicating that the person is a as the leader, the processor is to:
modify in real time, a manner in which a video stream of the person is presented on a graphical user interface (GUI);
wherein responsive to the comparison failing to indicate that the person is the leader, the processor is to:
apply a machine learning model to the multimodal content to identify a behavioral trait of the person, the behavioral trait comprising a facial feature of the person, a gesture made by the person or a speech pattern of the person;
determine, using the machine learning model, whether the person is the leader based on a comparison of the behavioral trait with behavioral profiles of individuals labeled as leaders; and
modify, based on the determination, the manner in which the video stream of the person is presented on the GUI; and
wherein the machine learning model is trained on training data comprising images of the individuals, audio recordings of the individuals and gesture data of the individuals.