US 11,817,118 B2
Enhancing review videos
Sourabh Gupta, Uttar Pradesh (IN); Mrinal Kumar Sharma, Jharkhand (IN); and Gourav Singhal, Delhi (IN)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by ADOBE INC., San Jose, CA (US)
Filed on Feb. 25, 2022, as Appl. No. 17/680,916.
Application 17/680,916 is a continuation of application No. 17/066,207, filed on Oct. 8, 2020, granted, now 11,302,360.
Prior Publication US 2022/0215860 A1, Jul. 7, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G11B 27/10 (2006.01); G06F 3/0482 (2013.01); G11B 27/06 (2006.01); G11B 27/031 (2006.01); G06F 40/279 (2020.01); G06F 40/30 (2020.01); G06Q 30/0282 (2023.01); G06Q 30/0601 (2023.01); G06N 20/00 (2019.01); G11B 27/11 (2006.01); G11B 27/34 (2006.01); G06F 40/216 (2020.01); G06F 40/284 (2020.01); G06F 17/00 (2019.01)
CPC G11B 27/102 (2013.01) [G06F 3/0482 (2013.01); G06F 40/279 (2020.01); G06F 40/30 (2020.01); G06N 20/00 (2019.01); G06Q 30/0282 (2013.01); G06Q 30/0623 (2013.01); G06Q 30/0641 (2013.01); G11B 27/031 (2013.01); G11B 27/06 (2013.01); G06F 40/216 (2020.01); G06F 40/284 (2020.01); G11B 27/11 (2013.01); G11B 27/34 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A computerized method comprising:
converting, by a text extraction module, audio from a review video for a product to text;
determining, by a topic identification module, topics in the text from the review video using keywords identified based at least in part on terms in text on a product page for the product, each topic corresponding to a time segment of the review video, wherein determining the topics in the text from the review video comprises:
dividing the review video into a plurality of time segments,
determining the topic for each time segment by: generating a feature vector from keywords in the text for the time segment of the review video, providing the feature vector to a classifier trained on labeled feature vectors, and receiving, from the classifier, a topic label determined by the classifier based on the feature vector, and
merging adjacent time segments having the same topic;
providing, by a user interface module, a user interface for presentation on a client device, the user interface including the review video and a user interface element for each topic;
receiving, by the user interface module, input indicative of a selection of a first user interface element identifying a first topic; and
causing, by a video playback module, playback of the review video at a time based on a first time segment corresponding to the first topic.