A Review on Machine and Deep Learning for Semiconductor Defect Classification in Scanning Electron Microscope Images

Rosa, Francisco López de la; Sánchez-Reolid, Roberto; Gómez-Sirvent, José L.; Morales, Rafael; Fernández-Caballero, Antonio

doi:10.3390/app11209508

Cited by 27 publications

(8 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the following article [21], authors used five machine learning models to classify minerals, including Logistic Regression, Linear Support Vector Machine, k-Nearest Neighbor, Random Forest, and Artificial Neuron Networks, and a deep learning CNN U-Net model, concluded that Random Forest outperformed with a F1 score of 0.92. In [22] authors classified the semiconductor defects in SEM images. It is evident that the prior research produced better outcomes than our own findings.…”

Section: Low Complexity Algorithm Resultsmentioning

confidence: 99%

Classification of FESEM Images of Nanostructures by Noval Deep Learning Algorithm

Patil,

Choudapur

et al. 2024

Preprint

View full text Add to dashboard Cite

SEM (Scanning Electron Microscopy) takes nanoscale pictures, whereas DL (Deep Learning) analyses data using neural networks. Image interpretation is streamlined by the collaboration of SEM and DL, which automates SEM image analysis and material characterisation. This integration improves productivity by quickly extracting relevant information from enormous datasets, highlighting subtle patterns that would otherwise be difficult to detect manually. Authors of this article consider 3 classes of SEM images of various magnifications and a total of 93 images is being produced for analysis and processing. Authors intend to enhance, categorize, and segment the pictures collected. SEM pictures are classified based on doping substance with the host material pure Cobalt Chromite, Neodymium doped Cobalt Chromite and Lanthanum doped Co-balt Chromite. Image quality and appearance are improved via augmentation techniques. The similarity and complexity of images in all three classes and the inclusion of images of different magnification posed a challenge in classification which hinder the accuracy rate of classification process. So authors use the statistical results of the tests to create semi-automatic method for classifying and labeling pictures generated by the SEM. Authors propose a low complexity algorithm, aimed to extract features and increase model performance. This approach offers efficiency gains by minimizing time and computational resources compared to pre-trained models, while maintaining consistent classification results. Achieved a training and testing accuracy of 100% and 78.26% respectively. SEM pictures are also classified using CNN and pre-trained models VGG16, Inception v3. To evaluate and compare performance, comparative studies are used to measure model correctness. SEM images are segmented into regions of interest using an integrated technique that combined the watershed and contours. A segmented picture is used to calculate the surface area of a Cobalt Chromite sample. The surface area of material is determined to be 41,02,628 nm 2 .

show abstract

Section: Low Complexity Algorithm Resultsmentioning

confidence: 99%

Classification of FESEM Images of Nanostructures by Noval Deep Learning Algorithm

Patil,

Choudapur

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Ref. 5 reviewed nine ML-based semiconductor SEM defect detection studies published in and before 2020. They found that Convolutional neural networks (CNNs) Ref.…”

Section: Related Workmentioning

confidence: 99%

Deep learning based defect classification and detection in SEM images: a mask R-CNN approach

Dey

Dehaerne

Halder

et al. 2022

Metrology, Inspection, and Process Control XXXVI

View full text Add to dashboard Cite

In this research work, we have demonstrated the application of Mask-RCNN (Regional Convolutional Neural Network), a deep-learning algorithm for computer vision and specifically object detection, to semiconductor defect inspection domain. Stochastic defect detection and classification during semiconductor manufacturing has grown to be a challenging task as we continuously shrink circuit pattern dimensions (e.g., for pitches less than 32 nm). Defect inspection and analysis by state-of-the-art optical and e-beam inspection tools is generally driven by some rule-based techniques, which in turn often causes to misclassification and thereby necessitating human expert intervention. In this work, we have revisited and extended our previous deep learning-based defect classification and detection method towards improved defect instance segmentation in SEM images with precise extent of defect as well as generating a mask for each defect category/instance. This also enables to extract and calibrate each segmented mask and quantify the pixels that make up each mask, which in turn enables us to count each categorical defect instances as well as to calculate the surface area in terms of pixels. We are aiming at detecting and segmenting different types of inter-class stochastic defect patterns such as bridge, break, and line collapse as well as to differentiate accurately between intra-class multi-categorical defect bridge scenarios (as thin/single/multi-line/horizontal/non-horizontal) for aggressive pitches as well as thin resists (High NA applications). Our proposed approach demonstrates its effectiveness both quantitatively and qualitatively.

show abstract

“…While these automated techniques have shown promise in detecting and analyzing defects, they still have some limitations in terms of robustness and performance when compared to the skill and expertise of a trained human operator. But in the recent years deep-learning enabled significant improvements [1].…”

Section: Introductionmentioning

confidence: 99%

A generic deep-learning-based defect segmentation model for electron micrographs for automatic defect inspection

Baderot

Hallal²,

Jacob³

et al. 2023

Metrology, Inspection, and Process Control XXXVII

View full text Add to dashboard Cite

Defect inspection is an important part in the semiconductor manufacturing. This task is tedious and time consuming if done manually. Therefore, reliably automating this task is a major challenge for many semiconductor manufacturers. In the recent years, deep-learning methods for object detection have demonstrated ever better performances. However, most of the publicly available models are trained on natural images and objects. Hence, most of them needs a long and data greedy training step to be used on industrial Transmission Electron Microscopes or Scanning Electron Microscope images. In this context, we propose a deep-learning based model to detect and segment defects in electron micrographs. Using SmartDef3 from Pollen Metrology, we annotated defects on images from several industrial applications. We split them in a training and validation dataset, with which an Instance segmentation model with state-of-the-art backbone is trained. The model is then evaluated on different use cases. Competitive performances on new data in terms of detection rates and segmentation quality are demonstrated and discussed. Furthermore, the model showed a relevant defect detection rate even on images that are not in the semiconductor domain, providing an interesting tool for defects detection on a new use case without new training data. This shows how deep learning strategies can help save time and costs by automation of defects inspections. Furthermore, advanced metrological analysis of the defect can be simultaneously obtained that help optimizing the manufacturing processes and reduce defect production rate.

show abstract

A Review on Machine and Deep Learning for Semiconductor Defect Classification in Scanning Electron Microscope Images

Cited by 27 publications

References 58 publications

Classification of FESEM Images of Nanostructures by Noval Deep Learning Algorithm

Classification of FESEM Images of Nanostructures by Noval Deep Learning Algorithm

Deep learning based defect classification and detection in SEM images: a mask R-CNN approach

A generic deep-learning-based defect segmentation model for electron micrographs for automatic defect inspection

Contact Info

Product

Resources

About