US 12,283,197 B2
Active learning for inspection tool
Nader Salman, Houston, TX (US); Victor Amblard, Paris (FR); and Vahagn Hakopian, Vienne (FR)
Assigned to SCHLUMBERGER TECHNOLOGY CORPORATION, Sugar Land, TX (US)
Appl. No. 17/597,520
Filed by Schlumberger Technology Corporation, Sugar Land, TX (US)
PCT Filed Jul. 10, 2020, PCT No. PCT/US2020/041597
§ 371(c)(1), (2) Date Jan. 10, 2022,
PCT Pub. No. WO2021/007514, PCT Pub. Date Jan. 14, 2021.
Claims priority of provisional application 62/872,609, filed on Jul. 1, 2019.
Claims priority of provisional application 62/966,156, filed on Jan. 27, 2020.
Prior Publication US 2022/0262104 A1, Aug. 18, 2022
Int. Cl. G06F 17/00 (2019.01); G06V 10/774 (2022.01); G06V 10/778 (2022.01); G06V 10/82 (2022.01); G06V 20/05 (2022.01); G09B 23/40 (2006.01); G01V 1/30 (2006.01)
CPC G09B 23/40 (2013.01) [G06V 10/7753 (2022.01); G06V 10/778 (2022.01); G06V 10/82 (2022.01); G06V 20/05 (2022.01); G01V 1/301 (2013.01)] 22 Claims
OG exemplary drawing
 
1. A method comprising:
receiving labeled images;
acquiring unlabeled images;
performing active learning by training an inspection learner using at least a portion of the labeled images to generate a trained inspection learner that, responsive to receipt of one of the unlabeled images by the trained inspection learner, outputs uncertainty information as to one or more features in the one of the unlabeled images;
generating, by a density learner, responsive to receipt of the one of the unlabeled images, a density metric as to the one or more features in the one of the unlabeled images, wherein generating the density metric comprises computing a mean similarity of an unlabeled image to other images;
training a typicality learner that generates a typicality metric responsive to receipt of an image;
computing a relevance score, a density score using the density metric, a typicality score using the typicality metric, and an uncertainty score using the uncertainty information;
based at least in part on at least one of the relevance score, the density score, the typicality score, or the uncertainty score, calling for analysis of one of the labeled images or for analysis of the one of the unlabeled images;
making a decision to call for labeling of the one of the unlabeled images based at least in part on the uncertainty information, the typicality metric, and the density metric;
receiving a label for the one of the unlabeled images; and
further training the inspection learner using the label.