US 12,306,872 B2
Metadata tag identification
Dingxian Wang, New Castle, WA (US); Hongxu Chen, Ryde (AU); Guandong Xu, Ryde (AU); and Li He, Beijing (CN)
Assigned to eBay Inc., San Jose, CA (US)
Filed by eBay Inc., San Jose, CA (US)
Filed on Mar. 18, 2024, as Appl. No. 18/607,949.
Application 18/607,949 is a continuation of application No. 17/500,455, filed on Oct. 13, 2021, granted, now 11,966,440.
Prior Publication US 2024/0220537 A1, Jul. 4, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/70 (2019.01); G06F 16/74 (2019.01); G06F 16/78 (2019.01); G06F 16/783 (2019.01); G06F 16/40 (2019.01); G06F 16/90 (2019.01)
CPC G06F 16/7867 (2019.01) [G06F 16/743 (2019.01); G06F 16/7834 (2019.01); G06F 16/7844 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
extracting content features from a video, the extracted content features including two or more of textual content features, image content features, and audio content features;
encoding, by an encoder of a neural network, the extracted content features to represent the extracted content features numerically as a vector;
decoding, by a decoder of the neural network that includes a plurality of gated recurrent units (GRUs) arranged in a sequence, the vector to identify content platform metadata tags indicative of content from the video by providing, as input to a first GRU in the sequence, an initialization state corresponding to a group of users for which the content platform metadata tags are to be generated; and
uploading the video to an online platform with the content platform metadata tags for presentation to the group of users.