Defect inspection system employing an image enhancing model and a defect detection model, and method of using the same

A method comprising:

• forming a hard mask layer over a target layer;
• forming a photoresist over the hard mask layer;
• patterning the photoresist using photolithography;
• etching the hard mask layer using the patterned photoresist as an etching mask to form a patterned hard mask;
• removing the photoresist;
• capturing a low resolution image of a region of the patterned hard mask with a camera;
• enhancing the low resolution image of the patterned hard mask with a pre-trained machine learning model to produce an enhanced resolution image of the patterned hard mask, wherein the pre-trained machine learning model is trained by iteratively performing:
• generating an upscaled image of a training low resolution image using a neural network,
• comparing the upscaled image to a training high resolution image corresponding to the training low resolution image to generate an error function,
• tuning parameters of the neural network based on a value of the error function,
• regenerating the upscaled image using parameters tuned by the tuning step,
• comparing the regenerated upscaled image to the training high resolution image to regenerate the error function, and
• storing the upscaled image as an enhanced resolution image in an enhanced resolution image database in response to determining the error function is below a pre-determined threshold;
• analyzing the enhanced resolution image of the patterned hard mas with a second machine learning model to determine whether the target layer has defects;
• and
• transferring the pattern of the patterned hard mask to the target layer.

An apparatus including a non-transitory machine-readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to perform operations comprising:

• controlling a processing tool to pattern a hard mask over a target layer;
• receiving, from a camera, a low resolution image of the hard mask;
• enhancing the low resolution image of the hard mask with a pre-trained machine learning model to produce an enhanced resolution image of the hard mask, wherein the pre-trained machine learning model is trained by iteratively performing:
• generating an upscaled image of a low resolution image using a neural network,
• comparing the upscaled image to a high resolution image corresponding to the low resolution image to generate an error function,
• tuning parameters of the neural network based on a value of the error function,
• regenerating the upscaled image using parameters tuned by the tuning, comparing the regenerated upscaled image to the high resolution image corresponding to the low resolution image to regenerate the error function, and
• storing the upscaled image as an enhanced resolution image in an enhanced resolution image database in response to determining the error function is below a pre-determined threshold;
• analyzing the enhanced resolution image of the hard mask with a second machine learning model to determine whether the target layer has defects;
• and
• controlling the processing tool to transfer the pattern of the hard mask to the target layer.