Cognex's Deep Learning tools can help manufacturers identify and classify true mold compound defects. This advanced vision solution uses a set of training images showing good and defective (NG) results, allowing the software to ignore anomalies within the margin of error and flag actual critical defects. Cognex's location tools can identify regions of interest (ROI). Once the ROI is defined, the defect detection tool identifies defects within that area. Then, the classification tool categorizes the different types of defects. With this information, production managers can not only improve yields of good ICs, but also use the classified data to diagnose and correct production issues, increasing profitability.