An apparatus for predicting a structure of a semiconductor device, the apparatus includes: at least one processor; a storage configured to store a learned model configured to predict the structure of the semiconductor device; and a memory configured to store at least one code, and at least one processor operatively connected to the memory and configured to execute the at least one code to: input non-destructive metrology data measured from the semiconductor device into the learned model, and predict the structure of the semiconductor device, based on the learned model, wherein the learned model is trained with training data including first data which is non-destructive metrology data and second data which is structural metrology data as reference data of the first data, and wherein the training data is refined based on a similarity of the training data in a space having a first axis corresponding to the first data and a second axis corresponding to the second data as reference axes.

Systems, methods, and media for determining a processing parameter associated with a lithography process. In some embodiments, image data of features on a substrate may be obtained, and the image data may be analyzed in Fourier space. Based on the analysis, an amplitude and a phase may be determined, and an overlay of the features may be determined based on the amplitude and the phase.

Autoencoder models may be used in the field of lithography to estimate, infer or predict a parameter of interest (e.g., metrology metrics). An autoencoder model is trained to predict a parameter by training it with measurement data (e.g., pupil images) of a substrate obtained from a measurement tool (e.g., optical metrology tool). Disclosed are methods and systems for synchronizing two or more autoencoder models for in-device metrology. Synchronizing two autoencoder models may configure the encoders of both autoencoder models to map from different signal spaces (e.g., measurement data obtained from different machines) to the same latent space, and the decoders to map from the same latent space to each autoencoder's respective signal space. Synchronizing may be performed for various purposes, including matching a measurement performance of one tool with another tool, and configuring a model to adapt to measurement process changes (e.g., changes in characteristics of the tool) over time.

A correction to an error of overlay measurement which accounts for target structure asymmetry using a neural network is described. According to embodiments, an overlay measurement accuracy can be improved by accounting for multiple and/or asymmetric perturbations in the target structure. A trained neural network is described which generates a correction value for overlay measurement based on a measure of asymmetry. Based on an as-measured overlay measurement, which may not account for target structure asymmetry, and the correction value, a true overlay measurement is determined-which can exhibit improved accuracy and reduced uncertainty versus uncorrected values.

Systems and methods for training a machine learning model for defect detection include obtaining training data including an inspection image of a fabricated integrated circuit (IC) and design layout data of the IC, and training a machine learning model using the training data. The machine learning model includes a first autoencoder and a second autoencoder. The first autoencoder includes a first encoder and a first decoder. The second autoencoder includes a second encoder and a second decoder. The second decoder is configured to obtain a first code outputted by the first encoder. The first decoder is configured to obtain a second code outputted by the second encoder.

A method for processing images for metrology using a charged particle beam tool may include obtaining, from the charged particle beam tool, an image of a portion of a sample. The method may further include processing the image using a first image processing module to generate a processed image. The method may further include determining image quality characteristics of the processed image and determining whether the image quality characteristics of the processed image satisfy predetermined imaging criteria. The method may further include in response to the image quality characteristics of the processed image not satisfying the imaging criteria, updating a tuning condition of the charged-particle beam tool, acquiring an image of the portion of the sample using the charged-particle beam tool that has the updated tuning condition, and processing the acquired image using the first image processing module to enable the processed acquired image to satisfy the predetermined imaging criteria.

The invention relates to a computer implemented method for simulating an aerial image of a model of a photolithography mask illuminated by incident electromagnetic waves, the method comprising: obtaining the model of the photolithography mask, the model describing the photolithography mask at least partially in a dimension orthogonal to the mask carrier plane; simulating the propagation of the incident electromagnetic waves through the model of the photolithography mask using a machine learning model, wherein the machine learning model maps the model of the photolithography mask to a representation of an electromagnetic field generated by the incident electromagnetic waves on the photolithography mask; obtaining the aerial image of the model of the photolithography mask by applying a simulation of an imaging process. The invention also relates to corresponding computer programs, computer-readable media and systems.

The current disclosure describes techniques for managing vertical alignment or overlay in semiconductor manufacturing using machine learning. Alignments of interconnection features in a fan-out WLP process are evaluated and managed through the disclosed techniques. Big data and machine learning are used to train a classification that correlates the overlay error source factors with overlay metrology categories. The overlay error source factors include tool signals. The trained classification includes a base classification and a Meta classification.

A system and methods for OCD metrology are provided including receiving training data for training an OCD machine learning (ML) model, the training data measured from multiple wafers and including multiple pairs of corresponding input and label datasets obtained from each respective wafer. The input dataset of each pair includes multiple scatterometric datasets, measured at multiple respective locations defined by a first map. The label dataset of each pair includes one or more critical dimension (CD) parameters of respective locations defined by a second map, the second map including at least one location not in the first map. The OCD ML model is then applied to a new set of scatterometric datasets, measured from locations of a new wafer, according to the first map, to generate predicted CD parameters of locations of the second map on the new wafer.

Disclosed is a method for determining a parameter of interest relating to at least one target on a substrate. The method comprises obtaining metrology data comprising at least one asymmetry signal, said at least one asymmetry signal comprising a difference or imbalance in a measurement parameter from the target; obtaining a trained model having been trained or configured to relate said at least one asymmetry signal to the parameter of interest, the trained model comprising at least one proxy for at least one nuisance component of the at least one asymmetry signal; and inferring said parameter of interest for said at least one target from said at least one asymmetry signal using the trained model.