Generate one kind of image (namely EUV or He10830 images which are not regularly available) from other kinds of images that are more routinely available.

We will predict the trend for SC 25 based on derived EUV irradiance variation patterns of the past ten SCs. We propose to generate monthly averaged F10.7 and EUV fluxes for the remaining SC 25 and the early phase of SC 26 (2025-2035). This figure demonstrates our initial success in the prediction of SOHO/SEM EUV irradiance using LSTM. Specifically, the model comprises a sequence input layer, an LSTM layer, a dropout layer, a fully-connected layer, and a linear activation function. An LSTM layer is a recurrent neural network (RNN) layer that learns long-term dependencies between time steps in time series and sequence data. SOHO/SEM irradiance in SC 23 and 24 are used for training. Based on the partially available data in the cycle 25, the prediction R2 for 26-34 nm irradiance (W/m2) and 0.1-50 nm irradiance (W/m2) is around 0.95.

Solar atmosphere modeling solves the partial differential equations (PDEs) for the magnetic induction equation, the Navier-Stokes equation, the continuity equation, and the energy conservation equation. These equations are discretized and solved on a grid, but the wide range of spatial and temporal scales and the complex nonlinear interactions between these scales require fine grids and high resolution, making these computations extremely expensive. The Fourier neural operator is used to accelerate solar atmosphere modeling.

A deep learning model is proposed to automatically segment and track the melt pool in the X-ray image sequence. The proposed deep learning structure uses spatiotemporal attention-aware features to automatically learn the spatiotemporal correlation in X-ray image sequence since melt pool changes smoothly in both spatial and temporal domains. By only training data from the spot melt printing pattern, our proposed method shows excellent extrapolation results when testing on data from the linear scan printing pattern.  

Differential equations are fundamental in modeling numerous physical systems, including thermal, manufacturing, and meteorological systems. Traditionally, numerical methods often approximate the solutions of complex systems modeled by differential equations. With the advent of modern deep learning, Physics-informed Neural Networks (PINNs) are evolving as a new paradigm for solving differential equations with a pseudo-closed form solution.

Unsupervised defect detection methods are applied to an unlabeled dataset by producing a ranked list based on defect scores. Unfortunately, many of the top-ranked instances by unsupervised algorithms are not defects, which leads to high false-positive rates. Active Defect Discovery (ADD) is proposed to overcome this deficiency, which sequentially selects instances to get the labeling information (defects or not). 

Laser powder bed fusion is a promising technology for local deposition and microstructure control, but it suffers from defects such as delamination and porosity due to the lack of understanding of melt pool dynamics. To study the fundamental behavior of the melt pool, both geometric and thermal sensing with high spatial and temporal resolutions are necessary. This work applies and integrates three advanced sensing technologies: synchrotron X-ray imaging, high-speed IR camera, and high-spatial-resolution IR camera to characterize the evolution of the melt pool shape, keyhole, vapor plume, and thermal evolution in Ti–6Al–4V and 410 stainless steel spot melt cases.

Abnormal states (with defects) occur much less frequently than normal ones (without defects) in a manufacturing process, the number of sensor data samples collected from a normal state is usually much more than that from an abnormal state. This issue causes imbalanced training data for classification analysis, thus deteriorating the performance of detecting abnormal states in the process. It is beneficial to generate effective artificial sample data for the abnormal states to make a more balanced training set.

Quality assurance is the main challenge in AM industry due to the possible time-varying processing conditions during AM process. Notably, new defects may occur during printing, which cannot be mitigated by offline analysis tools that focus on existing defects. This challenge motivates this work to develop online learning-based methods to deal with the new defects during printing.

The goal of this article is to develop a new super resolution method that integrates these different types of image stream data to improve both spatial and temporal resolutions, which is critical to obtaining more insightful information for more effective quality control of targeted processes or systems. 

Model compression is a crucial part of deploying neural networks (NNs), especially when the memory and storage of computing devices are limited in many applications. This paper focuses on two model compression techniques: low-rank approximation and weight pruning in neural networks, which are very popular nowadays. 

In real-world applications, the presence of missing pixels is a very common and challenging issue due to errors in the acquisition process or manufacturer defects. RPCA and RTPCA are not able to recover the background and foreground simultaneously with missing pixels. 

Robust PCA (RPCA) and its tensor extension, namely, Robust Tensor PCA (RTPCA), provide an effective framework for background/foreground separation by decomposing the data into low-rank and sparse components, which contain the background and the foreground (moving objects), respectively. RPCA and RTPCA are not able to separate the background, foreground, and noise simultaneously. As a result, the noise will contaminate the background or the foreground or both. There is a need to remove the noise from the background and foreground. To achieve the three component decomposition, a smooth sparse Robust Tensor Decomposition (SS-RTD) model is proposed to decompose the data into static background, smooth foreground, and noise, respectively.

A mask-guided attention (MGA) classifier, to improve the generalization and data efficiency of COVID-19 classification on lung CT scan images. The novelty of this method is compensating for the scarcity of data by employing more supervision with lesion masks, increasing the sensitivity of the model to COVID-19 manifestations, and helping both generalization and classification performance.