Abstract: An inspection system may develop an inspection recipe by generating N inspection images of a preliminary sample with one or more optical inspection sub-systems associated with N different optical inspection modes, generating probabilities that each of the locations of the preliminary sample are in background or defect classes using a classifier with the inspection images from at least some combinations of a number M of the optical inspection modes, where M is greater than one and less than N and corresponds to a number of the optical inspection modes to include in the inspection recipe, and selecting one of the combinations of M of the optical inspection modes based on a metric describing a distinction between the background and defect classes. The inspection system may further identify defects on a test sample using M inspection images generated with the selected combination of M of the optical inspection modes.