US 11,717,860 B2
Systems and methods for sorting of seeds
Mordekhay Shniberg, Magshimim (IL); Elad Carmon, Petach-Tikva (IL); Sarel Ashkenazy, Rehovot (IL); David Gedalyaho Vaisberger, Jerusalem (IL); and Sharon Ayal, Kibbutz Bet Nir (IL)
Assigned to SeedX Technolooles Inc., Wilmington, DE (US)
Filed by SeedX Technologies Inc., Wilmington, DE (US)
Filed on Dec. 20, 2022, as Appl. No. 18/84,624.
Application 18/084,624 is a continuation of application No. 16/769,273, granted, now 11,541,428, previously published as PCT/IB2018/059568, filed on Dec. 3, 2018.
Claims priority of provisional application 62/712,270, filed on Jul. 31, 2018.
Claims priority of provisional application 62/712,264, filed on Jul. 31, 2018.
Claims priority of provisional application 62/593,949, filed on Dec. 3, 2017.
Prior Publication US 2023/0121801 A1, Apr. 20, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. B07C 5/342 (2006.01); G06F 18/214 (2023.01); G06F 18/2321 (2023.01); G06F 18/2413 (2023.01); G06F 18/2431 (2023.01)
CPC B07C 5/3425 (2013.01) [G06F 18/214 (2023.01); G06F 18/2321 (2023.01); G06F 18/2413 (2023.01); G06F 18/2431 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A system for training at least one neural network for sorting of seeds, comprising:
at least one hardware processor executing a code, the code comprising:
accessing a training dataset comprising a plurality of training images of a plurality of sample seeds, each sample seed labelled with a ground truth label selected from a group consisting of: hybrid, and non-hybrid,
wherein for the group, visual features of the plurality of sample seeds are not explicitly defined; and
training the at least one neural network on the training dataset,
wherein the at least one neural network trained on the training dataset generates an outcome of hybrid or non-hybrid for each seed depicted in an input of at least one target image.