Classification of defect patterns of substrates

A system, comprising memory and a processing device coupled to the memory, wherein the processing device is to:

• obtain first data indicative of locations of defects of a first substrate comprising an inspection image of the first substrate;
• generate a first image indicating the locations of the defects of the first substrate, wherein generating the first image comprises:
• converting the inspection image to a frequency domain representation,
• applying one or more image sharpening filters in the frequency domain, and
• converting the frequency domain representation to a spatial domain to generate the first image;
• provide the first image indicating the locations of the defects of the first substrate to a trained machine learning model;
• obtain, as first output from the trained machine learning model, a classification of the locations of the defects of the first substrate;
• obtain a list of defect locations of a second substrate;
• generate a second image indicating locations of the defects of the second substrate based on the list of defect locations;
• obtain, as second output from the trained machine learning model, a classification of the locations of the defects of the second substrate;
• and
• perform a first corrective action in view of the first output from the trained machine learning model and a second corrective action in view of the second output from the trained machine learning model.

A non-transitory machine-readable storage medium storing instruction which, when executed, cause a processing device to perform operations comprising:

• obtaining, by a processing device, data indicative of locations of defects of a substrate;
• generating an image indicating the locations of the defects;
• providing the image indicating the locations of the defects to a trained machine learning model;
• obtaining, as output from the trained machine learning model, a classification of the locations of the defects;
• and
• performing a corrective action in view of the output from the trained machine learning model.