US 11,836,184 B2
Automatic metadata detector based on images
Vidhya Murali, Jersey City, NJ (US)
Assigned to Spotify AB, Stockholm (SE)
Filed by Spotify AB, Stockholm (SE)
Filed on Jan. 14, 2022, as Appl. No. 17/576,741.
Application 17/576,741 is a continuation of application No. 16/452,319, filed on Jun. 25, 2019, granted, now 11,263,493.
Prior Publication US 2022/0215211 A1, Jul. 7, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/55 (2019.01); G06F 16/45 (2019.01); G06N 3/08 (2023.01); G06F 16/438 (2019.01); G06N 3/04 (2023.01); G06F 18/24 (2023.01); G06F 18/214 (2023.01); G06F 18/21 (2023.01); G06V 10/764 (2022.01); G06V 10/776 (2022.01)
CPC G06F 16/55 (2019.01) [G06F 16/4387 (2019.01); G06F 16/45 (2019.01); G06F 18/214 (2023.01); G06F 18/217 (2023.01); G06F 18/24 (2023.01); G06N 3/04 (2013.01); G06N 3/08 (2013.01); G06V 10/764 (2022.01); G06V 10/776 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method for training a classifier to identify images belonging to a predetermined category, the method comprising:
receiving positive training images and negative training images;
detecting a first object in at least some of the positive training images and a second object in at least some of the negative training images; and
training the classifier to determine the first object as belonging to the predetermined category and the second object as not belonging to the predetermined category;
wherein, the classifier is trained to output a value indicative of a likelihood that an image belongs to the predetermined category.