US 12,112,517 B2
Diagnostic tool for deep learning similarity models
Oren Barkan, Tel Aviv (IL); Omri Armstrong, Tel Aviv (IL); Ori Katz, Tel Aviv (IL); and Noam Koenigstein, Tel Aviv (IL)
Assigned to Microsoft Technology Licensing, LLC., Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on Jul. 14, 2023, as Appl. No. 18/353,045.
Application 18/353,045 is a continuation of application No. 18/052,568, filed on Nov. 3, 2022, granted, now 11,769,315.
Application 18/052,568 is a continuation of application No. 17/084,468, filed on Oct. 29, 2020, granted, now 11,532,147, issued on Dec. 20, 2022.
Claims priority of provisional application 63/083,712, filed on Sep. 25, 2020.
Prior Publication US 2024/0029393 A1, Jan. 25, 2024
Int. Cl. G06V 10/46 (2022.01); G06F 18/2113 (2023.01); G06F 18/22 (2023.01); G06N 3/08 (2023.01); G06V 10/25 (2022.01)
CPC G06V 10/464 (2022.01) [G06F 18/2113 (2023.01); G06F 18/22 (2023.01); G06N 3/08 (2013.01); G06V 10/25 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
a processor;
a computer-readable medium storing instructions;
the instructions stored on the computer-readable medium executed by the processor to:
receive a test image;
determine, with an image classifier, an image classification of the test image;
determine, for the test image, a first activation map for at least one model layer using the determined image classification;
determine, for the test image, a first gradient map for the at least one model layer using the determined image classification; and
generate a first saliency map as a first element-wise function of the first activation map and the first gradient map.