%0 Figure %A Khater, Ismail M. %A Aroca-Ouellette, Stephane T. %A Meng, Fanrui %A Nabi, Ivan Robert %A Hamarneh, Ghassan %D 2019 %T Simulated dataset that contains two types of simulated blobs that are generated with known classes. %U https://plos.figshare.com/articles/figure/Simulated_dataset_that_contains_two_types_of_simulated_blobs_that_are_generated_with_known_classes_/9731930 %R 10.1371/journal.pone.0211659.g003 %2 https://plos.figshare.com/ndownloader/files/17428805 %K classification %K Caveolae %K transcript release factor %K molecule localization microscopy %K 3 D cluster %K PC 3-PTRF cells %K adaptor protein polymerase %K PC 3 prostate cancer cells %K scaffold %K approach %K method %K caveolae %K point clouds %K 10 PC 3 %K accuracy %K Cav 1 proteins %K plasma membrane invaginations %K PointNet model %K SMLM data %K CAVIN %K CNN %K super-resolution microscopy images %K molecule localization microscopy data %X

(A) The figure shows two samples from each class of the simulated data. The samples are taken from a dataset of 2000 blobs. The blue blobs are samples from the class 1 (positive) and the red blobs are samples from class 2 (negative). (B) The number of generated points per blob in both classes. It shows the percentage of generated blobs from both classes that have a variable number of points per blobs. There is a large overlap between many of blobs from both classes in terms of the number of generated points per blob.

%I PLOS ONE