US 12,267,752 B2
Sentiment capture by wireless network elements
Joseph Soryal, Glendale, NY (US); and Shawn Rajguru, Summit, NJ (US)
Assigned to AT&T Intellectual Property I, L.P., Atlanta, GA (US)
Filed by AT&T Intellectual Property I, L.P., Atlanta, GA (US)
Filed on Mar. 28, 2022, as Appl. No. 17/656,803.
Prior Publication US 2023/0308843 A1, Sep. 28, 2023
Int. Cl. H04W 4/30 (2018.01); G06V 10/771 (2022.01); G06V 40/16 (2022.01)
CPC H04W 4/30 (2018.02) [G06V 10/771 (2022.01); G06V 40/168 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
extracting, by a system comprising a processor, features of sensor data captured by a at least one sensor, communicatively coupled to the system via a wireless communication network and located in an area, wherein the sensor data is representative of respective persons present in the area, resulting in extracted features, wherein the extracted features include extracted properties of speech uttered by at least one person of the respective persons and extracted properties of video data,
wherein the extracted properties of the speech include: voice tones present in the speech, content of the speech, volume of the speech, and keywords that provide cues as to sentiment, and
wherein the extracted properties of the video data include: facial expressions of a depicted person included in the respective persons, a movement of the depicted person, and a posture of the depicted person;
determining, by the system, sentiment data, representative of an emotional condition of the respective persons present in the area, by correlating the extracted features to circumstantial properties associated with the area;
generating, by the system based on the sentiment data, a response to a query for information associated with the area, wherein the features of the sensor data are first features of first sensor data captured at a first time, wherein the extracted features are first extracted features, wherein the sentiment data is first sentiment data, and wherein the circumstantial properties comprise a first circumstantial property associated with the area at the first time;
extracting, by the system, second features of second sensor data captured by the at least one sensor at a second time, wherein the second sensor data is representative of the respective persons present in the area, resulting in second extracted features; and
determining, by the system, second sentiment data, representative of the emotional condition of the respective persons present in the area, by correlating the second extracted features to a second circumstantial property associated with the area at the second time.