view article

Figure 6
Output from the XChem CHiMP Detector Mask R-CNN model on members of the validation set from the XChem Detector Data Set. Blue bounding boxes denote objects in the `Drops' class and red bounding boxes denote objects in the `Crystals' class. Instance-segmentation mask colours are randomized. Green crosses denote candidate positions to dispense compound for fragment soaking (calculated via a distance transform). (a, d, g) Original micrographs of crystallization drops. (b, e, h) The corresponding bounding boxes and masks as predicted by the XChem CHiMP Detector network. Objects were predicted with a probability of ≥0.6. Candidate dispensing positions were calculated from output masks. (c, f, i) The ground-truth bounding boxes and masks from the XChem Detector Data Set as annotated by an expert crystallographer.

Journal logoSTRUCTURAL
BIOLOGY
ISSN: 2059-7983
Follow Acta Cryst. D
Sign up for e-alerts
Follow Acta Cryst. on Twitter
Follow us on facebook
Sign up for RSS feeds