The peculiar nature of whole slide imaging (WSI), digitizing conventional glass slides to obtain multiple high resolution images which capture microscopic details of a patient’s histopathological features, has garnered increased interest from the computer vision research community over the last two decades. Given the unique computational space and time complexity inherent to gigapixel-size whole slide image data, researchers have proposed novel machine learning algorithms to aid in the performance of diagnostic tasks in clinical pathology. One effective algorithm represents a Whole slide image as a bag of smaller image patches, which can be represented as low-dimension image patch embeddings. Weakly supervised deep-learning methods, such as cluster-constrained-attention multiple instance learning (CLAM), have shown promising results when combined with image patch embeddings. While traditional ensemble classifiers yield improved task performance, such methods come with a steep cost in model complexity. Through knowledge distillation, it is possible to retain some performance improvements from an ensemble, while minimizing costs to model complexity. In this work, we implement a weakly supervised ensemble using clustering-constrained-attention multiple-instance learners (CLAM), which uses attention and instance-level clustering to identify task salient regions and feature extraction in whole slides. By applying logit-based and attention-based knowledge distillation, we show it is possible to retain some performance improvements resulting from the ensemble at zero cost to model complexity.

Ptychography is a popular microscopic imaging modality for many scientific discoveries and sets the record for highest image resolution. Unfortunately, the high image resolution for ptychographic reconstruction requires significant amount of memory and computations, forcing many applications to compromise their image resolution in exchange for a smaller memory footprint and a shorter reconstruction time. In this paper, we propose a novel image gradient decomposition method that significantly reduces the memory footprint for ptychographic reconstruction by tessellating image gradients and diffraction measurements into tiles. In addition, we propose a parallel image gradient decomposition method that enables asynchronous point-to-point communications and parallel pipelining with minimal overhead on a large number of GPUs. Our experiments on a Titanate material dataset (PbTiO3) with 16632 probe locations show that our Gradient Decomposition algorithm reduces memory footprint by 51 times. In addition, it achieves time-to-solution within 2.2 minutes by scaling to 4158 GPUs with a super-linear strong scaling efficiency at 364% compared to runtimes at 6 GPUs. This performance is 2.7 times more memory efficient, 9 times more scalable and 86 times faster than the state-of-the-art algorithm.

Studies on domain adaptation (DA) for remote sensing (RS) imagery analysis lack consistency in selection and description of evaluation scenarios. Without properly characterizing datasets, model assumptions, and evaluation scenarios, it is difficult to objectively compare DA methods and reach conclusions about their suitability across different applications. With this motivation, this work seeks to empirically assess to which extent the interaction between data characteristics and model assumptions influence the effectiveness of DA methods. Using the widely explored task of building footprint segmentation as case study, we perform a large-scale study across over 200 domain adaptation scenarios that include variations across view angles, areas observed, and sensors used for data acquisition. Rather than adopting different model architectures or optimization criteria, we contrast the performances of two DA methods based on adversarial learning that differ only in their assumptions about source and target domains. Informed by metadata and data characteristics unveiled using traditional computer vision techniques as well as pre-trained deep models, we provide a detailed meta-analysis of experiments highlighting the importance of accurately considering data assumptions for DA in RS segmentation tasks. As a “cherry-picking” exercise demonstrates, different claims regarding which model is best could be made by selecting different subsets of evaluation scenarios. While well-calibrated assumptions can be beneficial, mismatching assumptions can lead to negative biases in DA applications. This study intends to motivate the community towards more consistent evaluation protocols, while providing recommendations and insights toward creating novel benchmark datasets, documenting data characteristics, application-specific knowledge, and model assumptions.

Semantic segmentation of images is an important computer vision task that emerges in a variety of application domains such as medical imaging, robotic vision and autonomous vehicles to name a few. While these domain-specific image analysis tasks involve relatively small image sizes (∼ 102 × 102), there are many applications that need to train machine learning models on image data with extents that are orders of magnitude larger (∼ 104 × 104). Training deep neural network (DNN) models on large extent images is extremely memory-intensive and often exceeds the memory limitations of a single graphical processing unit, a hardware accelerator of choice for computer vision workloads. Here, an efficient, sample parallel approach to train U-Net models on large extent image data sets is presented. Its advantages and limitations are analyzed and near-linear strong-scaling speedup demonstrated on 256 nodes (1536 GPUs) of the Summit supercomputer. Using a single node of the Summit supercomputer, an early evaluation of a recently released model parallel framework called GPipe is demonstrated to deliver ∼ 2X speedup in executing a U-Net model with an order of magnitude larger number of trainable parameters than reported before. Performance bottlenecks for pipelined training of U-Net models are identified and mitigation strategies to improve the speedups are discussed. Together, these results open up the possibility of combining both approaches into a unified scalable pipelined and data parallel algorithm to efficiently train U-Net models with very large receptive fields on data sets of ultra-large extent images.

There has recently been much work on the "wide limit" of neural networks, where Bayesian neural networks (BNNs) are shown to converge to a Gaussian process (GP) as all hidden layers are sent to infinite width. However, these results do not apply to architectures that require one or more of the hidden layers to remain narrow. In this paper, we consider the wide limit of BNNs where some hidden layers, called "bottlenecks", are held at finite width. The result is a composition of GPs that we term a "bottleneck neural network Gaussian process" (bottleneck NNGP). Although intuitive, the subtlety of the proof is in showing that the wide limit of a composition of networks is in fact the composition of the limiting GPs. We also analyze theoretically a single-bottleneck NNGP, finding that the bottleneck induces dependence between the outputs of a multi-output network that persists through extreme post-bottleneck depths, and prevents the kernel of the network from losing discriminative power at extreme post-bottleneck depths.

Significant strides have been made in supervised learning settings thanks to the successful application of deep learning. Now, recent work has brought the techniques of deep learning to bear on sequential decision processes in the area of deep reinforcement learning (DRL). Currently, little is known regarding hyperparameter optimization for DRL algorithms. Given that DRL algorithms are computationally intensive to train, and are known to be sample inefficient, optimizing model hyperparameters for DRL presents significant challenges to established techniques. We provide an open source, distributed Bayesian model-based optimization algorithm, HyperSpace, and show that it consistently outperforms standard hyperparameter optimization techniques across three DRL algorithms.

We developed a Nonlinear Level-set Learning (NLL) method for dimensionality reduction in high-dimensional function approximation with small data. This work is motivated by a variety of design tasks in real-world engineering applications, where practitioners would replace their computationally intensive physical models (e.g., high-resolution fluid simulators) with fast-to-evaluate predictive machine learning models, so as to accelerate the engineering design processes. There are two major challenges in constructing such predictive models: (a) high-dimensional inputs (e.g., many independent design parameters) and (b) small training data, generated by running extremely time-consuming simulations. Thus, reducing the input dimension is critical to alleviate the over-fitting issue caused by data insufficiency. Existing methods, including sliced inverse regression and active subspace approaches, reduce the input dimension by learning a linear coordinate transformation; our main contribution is to extend the transformation approach to a nonlinear regime. Specifically, we exploit reversible networks (RevNets) to learn nonlinear level sets of a high-dimensional function and parameterize its level sets in low-dimensional spaces. A new loss function was designed to utilize samples of the target functions' gradient to encourage the transformed function to be sensitive to only a few transformed coordinates. The NLL approach is demonstrated by applying it to three 2D functions and two 20D functions for showing the improved approximation accuracy with the use of nonlinear transformation, as well as to an 8D composite material design problem for optimizing the buckling-resistance performance of composite shells of rocket inter-stages.

We introduce a deep learning architecture, hierarchical self-attention networks (HiSANs), designed for classifying pathology reports and show how its unique architecture leads to a new state-of-the-art in accuracy, faster training, and clear interpretability. We evaluate performance on a corpus of 374,899 pathology reports obtained from the National Cancer Institute's (NCI) Surveillance, Epidemiology, and End Results (SEER) program. Each pathology report is associated with five clinical classification tasks - site, laterality, behavior, histology, and grade. We compare the performance of the HiSAN against other machine learning and deep learning approaches commonly used on medical text data - Naive Bayes, logistic regression, convolutional neural networks, and hierarchical attention networks (the previous state-of-the-art). We show that HiSANs are superior to other machine learning and deep learning text classifiers in both accuracy and macro F-score across all five classification tasks. Compared to the previous state-of-the-art, hierarchical attention networks, HiSANs not only are an order of magnitude faster to train, but also achieve about 1% better relative accuracy and 5% better relative macro F-score.

As machine learning models continue to increase in complexity, so does the potential number of free model parameters commonly known as hyperparameters. While there has been considerable progress toward finding optimal configurations of these hyperparameters, many optimization procedures are treated as black boxes. We believe optimization methods should not only return a set of optimized hyperparameters, but also give insight into the effects of model hyperparameter settings. To this end, we present HyperSpace, a parallel implementation of Bayesian sequential model-based optimization. HyperSpace leverages high performance computing (HPC) resources to better understand unknown, potentially non-convex hyperparameter search spaces. We show that it is possible to learn the dependencies between model hyperparameters through the optimization process. By partitioning large search spaces and running many optimization procedures in parallel, we also show that it is possible to discover families of good hyperparameter settings over a variety of models including unsupervised clustering, regression, and classification tasks.

In this paper, we present a Bayesian formulation of polynomial regression on a Riemannian manifold. Previous methods for fitting a curve to manifold-valued data have been formulated as geometric, least-squares estimation problems. We show that least-squares estimation on manifolds, much like the familiar Euclidean case, suffers from overfitting when using higher-order polynomials. Our Bayesian model mitigates this overfitting by placing a prior on the polynomial coefficients that shrinks their magnitude, analogous to Bayesian Euclidean regression with a Gaussian prior on the coefficients. We develop an algorithm for computing maximum a posteriori estimates of polynomial coefficients and the noise variance. Experiments on synthetically generated sphere data and a real shape regression problem demonstrate the advantages of our approach.