Deep learning for semantic segmentation of pattern

A computer program product comprising a non-transitory computer-readable medium having instructions, the instructions, when executed by a computer, configured to cause the computer to at least:

• obtain (i) training data comprising an input image of at least a part of a substrate having a plurality of features and comprising a truth image corresponding to the input image, (ii) a set of classes, each class corresponding to a feature of the plurality of features of the substrate within the input image, and (iii) a deep learning model;
• generate a predicted image by use of the deep learning model with the input image, wherein the deep learning model iteratively performs a convolution operation with at least one dilation kernel comprising a plurality of weights;
• assign a class of the set of classes to a feature within the predicted image based on matching of the feature with a corresponding feature within the truth image;
• and
• generate a trained deep learning model by iteratively assigning weights to the at least one dilated kernel based on a back propagation through the deep learning model using a loss function such that the predicted image reproduces the plurality of features of the input image.