An optical inspection is performed to detect potential defects within integrated circuit devices and a first electron-based inspection of less than all of the potential defects is performed to identify primary actual defects. A process window of manufacturing parameter settings used to manufacture the integrated circuit devices is identified and the integrated circuit devices manufactured using the manufacturing parameter settings inside the process window have less than a threshold number of the primary actual defects. To identify additional actual defects a second electron-based inspection is performed that is limited to selected ones of the potential defects in the integrated circuit devices that were manufactured using the manufacturing parameter settings inside the process window but were uninspected in the first electron-based inspection.

Described herein is a method for training a denoising model. The method includes obtaining a first set of simulated images based on design patterns. The simulated images may be clean and can be added with noise to generate noisy simulated images. The simulated clean and noisy images are used as training data to generate a denoising model.

The purpose of the present invention is to increase accuracy of a specific test using an electronic microscope and improve work efficiency. Provided is a system that identifies test recipe information corresponding to an object to be tested on the basis of attribute information about a testing sample, and analyzes and evaluates the object to be tested contained in the testing sample by checking image data and element analysis data that are acquired by a measuring device in accordance with a control program for the test recipe information, against reference image data and reference element analysis data that are used as evaluation references for the object to be tested.

A method can be used for sensing and processing image data for an object to be imaged. The object is scanned incompletely by virtue of regions (eB) of the object being sensed, where the sensed image regions (eB) alternate with non-sensed image regions (neB) of the object. Image data (rBD) of the non-sensed image regions (neB) are reconstructed on the basis of the sensed image data (eBD) of the sensed image regions (eB). A noise signal (N) of the sensed image data (eBD) of the sensed regions (eB) is ascertained and transferred to the reconstructed image data (rBD) of the non-sensed regions (neB), so that a user obtains a homogeneous visual impression in relation to the noise arising in the overall image data of the object visualized in a resultant overall image (rGB.sub.Inv).

A system for detecting defects includes a memory configured to store a program of instructions; and a processor configured execute the program of instructions to convert an SEM image into an image layout, determine a search space based on performing first layout matching for the image layout and a design layout, match the image layout and the design layout based on performing the second layout matching in the search space, and output defect information based on detecting defects the matched image layout and the matched design layout. The SEM image is an image obtained based on photographing a semiconductor pattern formed on a semiconductor wafer using the design layout based on using a scanning electron microscope.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for implementing a reservoir computing neural network. In one aspect there is provided a reservoir computing neural network comprising: (i) a brain emulation sub-network, and (ii) a prediction sub-network. The brain emulation sub-network is configured to process the network input in accordance with values of a plurality of brain emulation sub-network parameters to generate an alternative representation of the network input. The prediction sub-network is configured to process the alternative representation of the network input in accordance with values of a plurality of prediction sub-network parameters to generate the network output. The values of the brain emulation sub-network parameters are determined before the reservoir computing neural network is trained and are not adjusting during training of the reservoir computing neural network.

There is provided a method and a system configured obtain an image of a semiconductor specimen including one or more arrays, each including repetitive structural elements, and one or more regions, each region at least partially surrounding a corresponding array and including features different from the repetitive structural elements, wherein the PMC is configured to, during run-time scanning of the semiconductor specimen, perform a correlation analysis between pixel intensity of the image and pixel intensity of a reference image informative of at least one of the repetitive structural elements, to obtain a correlation matrix, use the correlation matrix to distinguish between one or more first areas of the image corresponding to the one or more arrays and one or more second areas of the image corresponding the one or more regions, and output data informative of the one or more first areas of the image.

An electron beam detection apparatus for a semiconductor device and an electron beam detection assembly are disclosed, the electron beam detection apparatus including a stage, which is configured to carry and hold the semiconductor device at a top surface of the stage, and is translatable in two directions orthogonal to each other, an aiming device, configured to determine a position of the semiconductor device in a coordinate system of the electron beam detection apparatus by capturing an image of the semiconductor device, the aiming device provided with a first field of view and a first optical axis, and an electron beam detection device, configured to detect an emergent electron beam exiting the semiconductor device by projecting an electron beam to the semiconductor device, the electron beam detection device provided with a second field of view and a second optical axis which is not consistent with the first optical axis.

A method is proposed for generating a series of ultra-thin sections of a microscopic sample (10), wherein the sections (11) are detached from the sample (10) using an ultramicrotome (100) and wherein the sections (11), which are detached from the sample (10) are caused to float on a liquid surface and are thereafter transferred onto a solid carrier element (20). For at least for some of the sections (11) detached from the sample (10) a position and an orientation on the solid carrier element (20) are determined by monitoring the placement of these sections (11) onto the solid carrier element (20) using a monitoring system (400) comprising a camera (410), obtaining monitoring data. A method (2000) for the three-dimensional reconstruction of a microscopic sample (10), a microtome system and a computer program are also part of the present invention.

In a learning phase, a processor of a sample observation device: stores design data on a sample in a storage resource; creates a first learning image as a plurality of input images; creates a second learning image as a target image; and learns a model related to image quality conversion with the first and second learning images. In a sample observation phase, the processor obtains, as an observation image, a second captured image output by inputting a first captured image obtained by imaging the sample with an imaging device to the model. The processor creates at least one of the first and second learning images based on the design data.