Jeffrey M. Ede

Jeffrey M. Ede

Compressed sensing can decrease scanning transmission electron microscopy electron dose and scan time with minimal information loss. Traditionally, sparse scans used in compressed sensing sample a static set of probing locations. However, dynamic scans that adapt to specimens are expected to be able to match or surpass the performance of static scans as static scans are a subset of possible dynamic scans. Thus, we present a prototype for a contiguous sparse scan system that piecewise adapts scan paths to specimens as they are scanned. Sampling directions for scan segments are chosen by a recurrent neural network based on previously observed scan segments. The recurrent neural network is trained by reinforcement learning to cooperate with a feedforward convolutional neural network that completes the sparse scans. This paper presents our learning policy, experiments, and example partial scans, and discusses future research directions. Source code, pretrained models, and training data is openly accessible at https://github.com/Jeffrey-Ede/adaptive-scans

Jeffrey M. Ede

Large, carefully partitioned datasets are essential to train neural networks and standardize performance benchmarks. As a result, we have set up new repositories to make our electron microscopy datasets available to the wider community. There are three main datasets containing 19769 scanning transmission electron micrographs, 17266 transmission electron micrographs, and 98340 simulated exit wavefunctions, and multiple variants of each dataset for different applications. To visualize image datasets, we trained variational autoencoders to encode data as 64-dimensional multivariate normal distributions, which we cluster in two dimensions by t-distributed stochastic neighbor embedding. In addition, we have improved dataset visualization with variational autoencoders by introducing encoding normalization and regularization, adding an image gradient loss, and extending t-distributed stochastic neighbor embedding to account for encoded standard deviations. Our datasets, source code, pretrained models, and interactive visualizations are openly available at https://github.com/Jeffrey-Ede/datasets.

Jeffrey M. Ede

Compressed sensing can increase resolution, and decrease electron dose and scan time of electron microscope point-scan systems with minimal information loss. Building on a history of successful deep learning applications in compressed sensing, we have developed a two-stage multiscale generative adversarial network to supersample scanning transmission electron micrographs with point-scan coverage reduced to 1/16, 1/25, ..., 1/100 px. We propose a novel non-adversarial learning policy to train a unified generator for multiple coverages and introduce an auxiliary network to homogenize prioritization of training data with varied signal-to-noise ratios. This achieves root mean square errors of 3.23% and 4.54% at 1/16 px and 1/100 px coverage, respectively; within 1% of errors for networks trained for each coverage individually. Detailed error distributions are presented for unified and individual coverage generators, including errors per output pixel. In addition, we present a baseline one-stage network for a single coverage and investigate numerical precision for web serving. Source code, training data, and pretrained models are publicly available at https://github.com/Jeffrey-Ede/DLSS-STEM

Jeffrey M. Ede, Richard Beanland

Artificial neural network training with stochastic gradient descent can be destabilized by "bad batches" with high losses. This is often problematic for training with small batch sizes, high order loss functions or unstably high learning rates. To stabilize learning, we have developed adaptive learning rate clipping (ALRC) to limit backpropagated losses to a number of standard deviations above their running means. ALRC is designed to complement existing learning algorithms: Our algorithm is computationally inexpensive, can be applied to any loss function or batch size, is robust to hyperparameter choices and does not affect backpropagated gradient distributions. Experiments with CIFAR-10 supersampling show that ALCR decreases errors for unstable mean quartic error training while stable mean squared error training is unaffected. We also show that ALRC decreases unstable mean squared errors for partial scanning transmission electron micrograph completion. Our source code is publicly available at https://github.com/Jeffrey-Ede/ALRC

Jeffrey M. Ede, Richard Beanland

Compressed sensing algorithms are used to decrease electron microscope scan time and electron beam exposure with minimal information loss. Following successful applications of deep learning to compressed sensing, we have developed a two-stage multiscale generative adversarial neural network to complete realistic 512$\times$512 scanning transmission electron micrographs from spiral, jittered gridlike, and other partial scans. For spiral scans and mean squared error based pre-training, this enables electron beam coverage to be decreased by 17.9$\times$ with a 3.8\% test set root mean squared intensity error, and by 87.0$\times$ with a 6.2\% error. Our generator networks are trained on partial scans created from a new dataset of 16227 scanning transmission electron micrographs. High performance is achieved with adaptive learning rate clipping of loss spikes and an auxiliary trainer network. Our source code, new dataset, and pre-trained models have been made publicly available at https://github.com/Jeffrey-Ede/partial-STEM

Jeffrey M. Ede

We present 14 autoencoders, 15 kernels and 14 multilayer perceptrons for electron micrograph restoration and compression. These have been trained for transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM) and for both (TEM+STEM). TEM autoencoders have been trained for 1$\times$, 4$\times$, 16$\times$ and 64$\times$ compression, STEM autoencoders for 1$\times$, 4$\times$ and 16$\times$ compression and TEM+STEM autoencoders for 1$\times$, 2$\times$, 4$\times$, 8$\times$, 16$\times$, 32$\times$ and 64$\times$ compression. Kernels and multilayer perceptrons have been trained to approximate the denoising effect of the 4$\times$ compression autoencoders. Kernels for input sizes of 3, 5, 7, 11 and 15 have been fitted for TEM, STEM and TEM+STEM. TEM multilayer perceptrons have been trained with 1 hidden layer for input sizes of 3, 5 and 7 and with 2 hidden layers for input sizes of 5 and 7. STEM multilayer perceptrons have been trained with 1 hidden layer for input sizes of 3, 5 and 7. TEM+STEM multilayer perceptrons have been trained with 1 hidden layer for input sizes of 3, 5, 7 and 11 and with 2 hidden layers for input sizes of 3 and 7. Our code, example usage and pre-trained models are available at https://github.com/Jeffrey-Ede/Denoising-Kernels-MLPs-Autoencoders

Jeffrey M. Ede

We present an atrous convolutional encoder-decoder trained to denoise 512$\times$512 crops from electron micrographs. It consists of a modified Xception backbone, atrous convoltional spatial pyramid pooling module and a multi-stage decoder. Our neural network was trained end-to-end to remove Poisson noise applied to low-dose ($\ll$ 300 counts ppx) micrographs created from a new dataset of 17267 2048$\times$2048 high-dose ($>$ 2500 counts ppx) micrographs and then fine-tuned for ordinary doses (200-2500 counts ppx). Its performance is benchmarked against bilateral, non-local means, total variation, wavelet, Wiener and other restoration methods with their default parameters. Our network outperforms their best mean squared error and structural similarity index performances by 24.6% and 9.6% for low doses and by 43.7% and 5.5% for ordinary doses. In both cases, our network's mean squared error has the lowest variance. Source code and links to our new high-quality dataset and trained network have been made publicly available at https://github.com/Jeffrey-Ede/Electron-Micrograph-Denoiser

Jeffrey M. Ede

This doctoral thesis covers some of my advances in electron microscopy with deep learning. Highlights include a comprehensive review of deep learning in electron microscopy; large new electron microscopy datasets for machine learning, dataset search engines based on variational autoencoders, and automatic data clustering by t-distributed stochastic neighbour embedding; adaptive learning rate clipping to stabilize learning; generative adversarial networks for compressed sensing with spiral, uniformly spaced and other fixed sparse scan paths; recurrent neural networks trained to piecewise adapt sparse scan paths to specimens by reinforcement learning; improving signal-to-noise; and conditional generative adversarial networks for exit wavefunction reconstruction from single transmission electron micrographs. This thesis adds to my publications by presenting their relationships, reflections, and holistic conclusions. This version of my thesis is typeset for online dissemination to improve readability, whereas the thesis submitted to the University of Warwick in support of my application for the degree of Doctor of Philosophy in Physics is typeset for physical printing and binding.

Jeffrey M. Ede

Deep learning is transforming most areas of science and technology, including electron microscopy. This review paper offers a practical perspective aimed at developers with limited familiarity. For context, we review popular applications of deep learning in electron microscopy. Afterwards, we discuss hardware and software needed to get started with deep learning and interface with electron microscopes. We then review neural network components, popular architectures, and their optimization. Finally, we discuss future directions of deep learning in electron microscopy.