Methods and systems for training a machine learning based measurement model conditioned by at least one regularization control parameter are described herein. A ML based measurement model conditioned by at least one regularization control parameter is trained for different control parameter values. A regularization control value provided as input to the trained ML based measurement model defines the regularization condition at inference. In a further aspect, an optimal value of a regularization control parameter is selected based on measurement performance on a set of measurement data. As measurement conditions change, the optimal value is reevaluated based on measurement performance on an updated set of measurement data that reflects the changing measurement conditions. In another further aspect, changes in measurement conditions and reevaluation of a regularization control value are performed automatically as measurement data is collected by a measurement system without interruption of the measurement process.

Disclosed is a program code and a non-transitory computer readable medium including the program code, in which the program code, when executed by a processor, causes an apparatus including the processor to perform operations of selecting a plurality of target patterns from a mask layout, generating an aerial image based on a source system including a plurality of point sources and the mask layout, constructing an objective function based on a plurality of NILS values corresponding to the plurality of target patterns in the aerial image, optimizing the source system such that the objective function has a maximum value, and outputting an optimized source system, and the optimized source system includes a combination of a plurality of effective factors corresponding to the plurality of point sources.

A system, including: a charged-particle beam inspection apparatus configured to scan a sample that includes a target with a plurality of pattern layers; and a controller including circuitry, configured to: obtain detection data in response to a scan of the target; and determine one or more characteristics of the sample in dependence on the obtained detection data and a model, wherein, for each of the plurality of pattern layers of the target, the model has a term that is dependent on the properties of the pattern layer.

Disclosed is a method of determining an illumination-detection system alignment of an illumination-detection system describing alignment of at least one detector and/or measurement illumination of a metrology apparatus in terms of two or more illumination-detection system alignment parameters, each illumination-detection system alignment parameter relating to a respective degree of freedom for aligning the detector and/or the measurement illumination. The method comprises obtaining a diffraction pattern relating to diffraction of broadband radiation from a structure; transforming each of one or more diffraction orders of the diffraction pattern to a respective region coordinate system, each region coordinate system comprising a first axis and a second axis, each region coordinate system being such that said first axis is aligned in relation to a direction of an intensity metric of each transformed diffraction order; and determining illumination-detection system alignment parameter values for the illumination-detection system alignment parameters.

A method to determine a performance indicator indicative of alignment performance of a processed substrate. The method includes obtaining measurement data including a plurality of measured position values of alignment marks on the substrate and calculating a positional deviation between each measured position value and a respective expected position value. These positional deviations are used to determine a directional derivative between the alignment marks, and the directional derivatives are used to determine at least one directional derivative performance indicator.

Various examples herein describe an apparatus and related method to calibrate two in-line lithographic tools used as part of in-line, panel-production equipment. The two lithographic tools are used to expose opposing sides of a panel or substrate, such as, for example, an Advanced Integrated Circuit Substrates (AICS) panel or another type of panel or substrate, such as a copper-clad laminate (CCL) panel). The panel may be, for example, a flat-panel display or another substrate type. Other systems, apparatuses, and methods are also disclosed.

Systems and methods for reducing prediction uncertainty in a prediction model associated with a patterning process are described. These may be used in calibrating a process model associated with the patterning process, for example. Reducing the uncertainty in the prediction model may include determining a prediction uncertainty parameter based on prediction data. The prediction data may be determined using the prediction model. The prediction model may have been calibrated with calibration data. The prediction uncertainty parameter may be associated with variation in the prediction data. Reducing the uncertainty in the prediction model may include selecting a subset of process data based on the prediction uncertainty parameter; and recalibrating the prediction model using the calibration data and the selected subset of the process data.

An illumination compensation method, comprising: acquiring working gaps between an imaging film stack and a mask in an exposure imaging system; calculating imaging light field intensities of a photoresist layer in the imaging film stack corresponding to the different working gaps, and, constructing a functional dependence of imaged pattern critical dimension in the photoresist layer of the imaging film stack on the working gap based on said imaging light field intensity critical dimension; calculating compensation parameters required for illumination light according to the said functional dependence of imaged pattern critical dimension on working gap; and, performing illumination compensation on the exposure imaging system according to the compensation parameters.

A contour extraction model learning device for detecting a contour of a semiconductor lithography pattern includes a memory storing a contour extraction training program, and a processor configured to execute the contour extraction training program stored in the memory, wherein the contour extraction training program extracts a first contour image by inputting a SEM image of a new pattern to a contour extraction unit, generates a virtual SEM image by inputting the first contour image to a style transfer model, and trains the contour extraction model based on a training dataset in which the first contour image is matched with the virtual SEM image.

A method for generating a retargeted pattern for a target pattern to be printed on a substrate. The method includes obtaining (i) the target pattern comprising at least one feature, the at least one feature having geometry including a first dimension and a second dimension, and (ii) a plurality of biasing rules defined as a function of the first dimension, the second dimension, and a property associated with features of the target pattern within a measurement region; determining values of the property at a plurality of locations on the at least one feature of the target pattern, each location surrounded by the measurement region; selecting, from the plurality of biasing rules based on the values of the property, a sub-set of biases; and generating the retargeted pattern by applying the selected sub-set of biases to the at least one feature of the target pattern.