US 12,406,031 B1
Generating predictions characterizing images using glimpses extracted from the images
Volodymyr Mnih, Toronto (CA); and Koray Kavukcuoglu, London (GB)
Assigned to GDM Holding LLC, Mountain View, CA (US)
Filed by DeepMind Technologies Limited, London (GB)
Filed on Feb. 22, 2024, as Appl. No. 18/584,717.
Application 18/584,717 is a continuation of application No. 17/737,544, filed on May 5, 2022, granted, now 11,941,088.
Application 17/737,544 is a continuation of application No. 16/927,159, filed on Jul. 13, 2020, granted, now 11,354,548, issued on Jun. 7, 2022.
Application 16/927,159 is a continuation of application No. 16/250,320, filed on Jan. 17, 2019, granted, now 10,748,041, issued on Aug. 18, 2020.
Application 16/250,320 is a continuation of application No. 14/731,348, filed on Jun. 4, 2015, granted, now 10,223,617, issued on Mar. 5, 2019.
Claims priority of provisional application 62/009,097, filed on Jun. 6, 2014.
Int. Cl. G06V 10/44 (2022.01); G06F 18/2431 (2023.01); G06V 20/80 (2022.01); G06V 30/194 (2022.01); G06V 30/413 (2022.01)
CPC G06F 18/2431 (2023.01) [G06V 10/44 (2022.01); G06V 20/80 (2022.01); G06V 30/194 (2022.01); G06V 30/413 (2022.01)] 20 Claims
OG exemplary drawing
 
10. A system comprising:
one or more computers; and
one or more storage devices communicatively coupled to the one or more computers, wherein the one or more storage devices store instructions that, when executed by the one or more computers, cause the one or more computers to perform operations comprising:
generating a prediction characterizing one or more images, comprising:
extracting a plurality of glimpses from the one or more images, comprising, at each iteration preceding a last iteration of a plurality of iterations:
receiving data identifying a current location in a current image of the one or more images;
extracting a current glimpse from the current image at the current location;
processing a network input derived from the current glimpse extracted from the current image at the current location using a location neural network, in accordance with values of a set of location neural network parameters, to generate a next location in a next image of the one or more images; and
providing the next location in the next image for processing at a next iteration of the plurality of iterations; and
generating the prediction characterizing the one or more images using one or more of the plurality of glimpses extracted from the one or more images.