Recent Advances In Hardware Research Articles

Peptides in chemical space

Peptides, defined as sequences of amino acids up to approximately 50 residues in length, represent an extremely large reservoir of potentially bioactive compounds, referred to here as the peptide chemical space. Recent advances in computer hardware and software have led to a wide application of computational methods to explore this chemical space. Here, we review different in silico approaches including structure-based design, genetic algorithms, and machine learning. We also review the use of molecular fingerprints to sample virtual libraries and to visualize the peptide chemical space. Finally, we present an overview of the known peptide chemical space in form of an interactive map representing 40,531 peptides collected from eleven open-access peptide and peptide-containing databases, accessible at https://tm.gdb.tools/map4/peptide_databases_tmap/. These peptides are displayed as TMAP (Tree-Map) according to their molecular fingerprint similarity computed using MAP4, a MinHashed atom pair fingerprint well suited to analyze large molecules.

Market Graph Clustering via QUBO and Digital Annealing

We present a novel technique for cardinality-constrained index-tracking, a common task in the financial industry. Our approach is based on market graph models. We model our reference indices as market graphs and express the index-tracking problem as a quadratic K-medoids clustering problem. We take advantage of a purpose-built hardware architecture to circumvent the NP-hard nature of the problem and solve our formulation efficiently. The main contributions of this article are bridging three separate areas of the literature, market graph models, K-medoid clustering and quadratic binary optimization modeling, to formulate the index-tracking problem as a binary quadratic K-medoid graph-clustering problem. Our initial results show we accurately replicate the returns of various market indices, using only a small subset of their constituent assets. Moreover, our binary quadratic formulation allows us to take advantage of recent hardware advances to overcome the NP-hard nature of the problem and obtain solutions faster than with traditional architectures and solvers.

Lattice: A Vision for Machine Learning, Data Engineering, and Policy Considerations for Digital Agriculture at Scale

Digital agriculture, with the incorporation of Internet-of-Things (IoT)-based technologies, presents the ability to control a system at multiple levels (individual, local, regional, and global) and generates tools that allow for improved decision making and higher productivity. Recent advances in IoT hardware, e.g., networks of heterogeneous embedded devices, and software, e.g., lightweight computer vision algorithms and cloud optimization solutions, make it possible to efficiently process data from diverse sources in a connected (smart) farm. By interconnecting these IoT devices, often across large geographical distances, it is possible to collect data at different time scales, including in near real-time (i.e., with delays of only a few tens of seconds). This data can then be used for actionable insights, e.g., precise applications of soil supplements and reduced environmental footprint. Through LATTICE, we present an integrated vision for IoT solutions, data processing, and actionable analytics for digital agriculture. We couple this with discussion of economics and policy considerations that will underlie adoption of such IoT and ML technologies. Our paper starts off with the types of datasets in typical field operations, followed by the lifecycle for the data and storage, cloud and edge analytics, and fast information-retrieval solutions. We discuss what algorithms are proving to be most impactful in this space, e.g., approximate data analytics and on-device/in-network processing. We conclude by discussing analytics for alternative agriculture for generation of biofuels and policy challenges in the implementation of digital agriculture in the wild.

Special Section on Full-Duplex Transceivers for Future Networks: Theory and Techniques

Conventional wireless communication systems operate in a half-duplex mode, i.e., current radios cannot transmit and receive at the same time and on the same frequency. Full-duplex wireless operation was generally assumed to be impossible due to the great difference in transmit and receive signal power levels. However, recent advances in antenna, hardware, and signal processing techniques have shown that full-duplex operation is practically feasible. Thanks to novel combinations of antenna, analog, and digital cancellation techniques, self-interference suppression of 80–110 dB can be made possible. The feasibility in building a practical full-duplex radio using off-the-shelf hardware and software-defined radios therefore alleviates many problems in wireless network design.

Applications of Deep Learning in Molecule Generation and Molecular Property Prediction.

Recent advances in computer hardware and software have led to a revolution in deep neural networks that has impacted fields ranging from language translation to computer vision. Deep learning has also impacted a number of areas in drug discovery, including the analysis of cellular images and the design of novel routes for the synthesis of organic molecules. While work in these areas has been impactful, a complete review of the applications of deep learning in drug discovery would be beyond the scope of a single Account. In this Account, we will focus on two key areas where deep learning has impacted molecular design: the prediction of molecular properties and the de novo generation of suggestions for new molecules.One of the most significant advances in the development of quantitative structure-activity relationships (QSARs) has come from the application of deep learning methods to the prediction of the biological activity and physical properties of molecules in drug discovery programs. Rather than employing the expert-derived chemical features typically used to build predictive models, researchers are now using deep learning to develop novel molecular representations. These representations, coupled with the ability of deep neural networks to uncover complex, nonlinear relationships, have led to state-of-the-art performance. While deep learning has changed the way that many researchers approach QSARs, it is not a panacea. As with any other machine learning task, the design of predictive models is dependent on the quality, quantity, and relevance of available data. Seemingly fundamental issues, such as optimal methods for creating a training set, are still open questions for the field. Another critical area that is still the subject of multiple research efforts is the development of methods for assessing the confidence in a model.Deep learning has also contributed to a renaissance in the application of de novo molecule generation. Rather than relying on manually defined heuristics, deep learning methods learn to generate new molecules based on sets of existing molecules. Techniques that were originally developed for areas such as image generation and language translation have been adapted to the generation of molecules. These deep learning methods have been coupled with the predictive models described above and are being used to generate new molecules with specific predicted biological activity profiles. While these generative algorithms appear promising, there have been only a few reports on the synthesis and testing of molecules based on designs proposed by generative models. The evaluation of the diversity, quality, and ultimate value of molecules produced by generative models is still an open question. While the field has produced a number of benchmarks, it has yet to agree on how one should ultimately assess molecules "invented" by an algorithm.

Automated analysis of foraminifera fossil records by image classification using a convolutional neural network

Abstract. Manual identification of foraminiferal morphospecies or morphotypes under stereo microscopes is time consuming for micropalaeontologists and not possible for nonspecialists. Therefore, a long-term goal has been to automate this process to improve its efficiency and repeatability. Recent advances in computation hardware have seen deep convolutional neural networks emerge as the state-of-the-art technique for image-based automated classification. Here, we describe a method for classifying large foraminifera image sets using convolutional neural networks. Construction of the classifier is demonstrated on the publicly available Endless Forams image set with a best accuracy of approximately 90 %. A complete automatic analysis is performed for benthic species dated to the last deglacial period for a sediment core from the north-eastern Pacific and for planktonic species dated from the present until 180 000 years ago in a core from the western Pacific warm pool. The relative abundances from automatic counting based on more than 500 000 images compare favourably with manual counting, showing the same signal dynamics. Our workflow opens the way to automated palaeoceanographic reconstruction based on computer image analysis and is freely available for use.

Online Path Sampling Control with Progressive Spatio-temporal Filtering

This work introduces a novel technique called progressive spatio-temporal filtering, an efficient method to build all-frequency approximations to the light transport distribution into a scene by filtering individual samples produced by an underlying path sampler, using online, iterative algorithms and data-structures that exploit both the spatial and temporal coherence of the approximated light field. Unlike previous approaches, the proposed method is both more efficient, due to its use of an iterative temporal feedback loop that massively improves convergence to a noise-free approximant, and more flexible, due to its introduction of a spatio-directional hashing representation that allows to encode directional variations like those due to glossy reflections. We then introduce four different methods to employ the resulting approximations to control the underlying path sampler and/or modify its associated estimator, greatly reducing its variance and enhancing its robustness to complex lighting scenarios. The core algorithms are highly scalable and low-overhead, requiring only minor modifications to an existing path tracer. Combined with the most recent advances in ray-tracing hardware and denoising algorithms, our work enables performing high-quality light transport simulations including complex visibility and transport phenomena in real time.

Generating Elevation Surface from a Single RGB Remotely Sensed Image Using Deep Learning

Generating Digital Elevation Models (DEM) from satellite imagery or other data sources constitutes an essential tool for a plethora of applications and disciplines, ranging from 3D flight planning and simulation, autonomous driving and satellite navigation, such as GPS, to modeling water flow, precision farming and forestry. The task of extracting this 3D geometry from a given surface hitherto requires a combination of appropriately collected corresponding samples and/or specialized equipment, as inferring the elevation from single image data is out of reach for contemporary approaches. On the other hand, Artificial Intelligence (AI) and Machine Learning (ML) algorithms have experienced unprecedented growth in recent years as they can extrapolate rules in a data-driven manner and retrieve convoluted, nonlinear one-to-one mappings, such as an approximate mapping from satellite imagery to DEMs. Therefore, we propose an end-to-end Deep Learning (DL) approach to construct this mapping and to generate an absolute or relative point cloud estimation of a DEM given a single RGB satellite (Sentinel-2 imagery in this work) or drone image. The model has been readily extended to incorporate available information from the non-visible electromagnetic spectrum. Unlike existing methods, we only exploit one image for the production of the elevation data, rendering our approach less restrictive and constrained, but suboptimal compared to them at the same time. Moreover, recent advances in software and hardware allow us to make the inference and the generation extremely fast, even on moderate hardware. We deploy Conditional Generative Adversarial networks (CGAN), which are the state-of-the-art approach to image-to-image translation. We expect our work to serve as a springboard for further development in this field and to foster the integration of such methods in the process of generating, updating and analyzing DEMs.

Moving analytical ultracentrifugation software to a good manufacturing practices (GMP) environment

Recent advances in instrumentation have moved analytical ultracentrifugation (AUC) closer to a possible validation in a Good Manufacturing Practices (GMP) environment. In order for AUC to be validated for a GMP environment, stringent requirements need to be satisfied; analysis procedures must be evaluated for consistency and reproducibility, and GMP capable data acquisition software needs to be developed and validated. These requirements extend to multiple regulatory aspects, covering documentation of instrument hardware functionality, data handling and software for data acquisition and data analysis, process control, audit trails and automation. Here we review the requirements for GMP validation of data acquisition software and illustrate software solutions based on UltraScan that address these requirements as far as they relate to the operation and data handling in conjunction with the latest analytical ultracentrifuge, the Optima AUC by Beckman Coulter. The software targets the needs of regulatory agencies, where AUC plays a critical role in the solution-based characterization of biopolymers and macromolecular assemblies. Biopharmaceutical and regulatory agencies rely heavily on this technique for characterizations of pharmaceutical formulations, biosimilars, injectables, nanoparticles, and other soluble therapeutics. Because of its resolving power, AUC is a favorite application, despite the current lack of GMP validation. We believe that recent advances in standards, hardware, and software presented in this work manage to bridge this gap and allow AUC to be routinely used in a GMP environment. AUC has great potential to provide more detailed information, at higher resolution, and with greater confidence than other analytical techniques, and our software satisfies an urgent need for AUC operation in the GMP environment. The software, including documentation, are publicly available for free download from Github. The multi-platform software is licensed by the LGPL v.3 open source license and supports Windows, Mac and Linux platforms. Installation instructions and a mailing list are available from ultrascan.aucsolutions.com.

Motion correction in magnetic resonance spectroscopy.

In vivo proton magnetic resonance spectroscopy and spectroscopic imaging (MRS/MRSI) are valuable tools to study normal and abnormal human brain physiology. However, they are sensitive to motion, due to strong crusher gradients, long acquisition times, reliance on high magnetic field homogeneity, and particular acquisition methods such as spectral editing. The effects of motion include incorrect spatial localization, phase fluctuations, incoherent averaging, line broadening, and ultimately quantitation errors. Several retrospective methods have been proposed to correct motion-related artifacts. Recent advances in hardware also allow prospective (real-time) correction of the effects of motion, including adjusting voxel location, center frequency, and magnetic field homogeneity. This article reviews prospective and retrospective methods available in the literature and their implications for clinical MRS/MRSI. In combination, these methods can attenuate or eliminate most motion-related artifacts and facilitate the acquisition of high-quality data in the clinical research setting.

Energy and Policy Considerations for Modern Deep Learning Research

The field of artificial intelligence has experienced a dramatic methodological shift towards large neural networks trained on plentiful data. This shift has been fueled by recent advances in hardware and techniques enabling remarkable levels of computation, resulting in impressive advances in AI across many applications. However, the massive computation required to obtain these exciting results is costly both financially, due to the price of specialized hardware and electricity or cloud compute time, and to the environment, as a result of non-renewable energy used to fuel modern tensor processing hardware. In a paper published this year at ACL, we brought this issue to the attention of NLP researchers by quantifying the approximate financial and environmental costs of training and tuning neural network models for NLP (Strubell, Ganesh, and McCallum 2019). In this extended abstract, we briefly summarize our findings in NLP, incorporating updated estimates and broader information from recent related publications, and provide actionable recommendations to reduce costs and improve equity in the machine learning and artificial intelligence community.

Development of basic building blocks for cryo-EM: the emcore and emvis software libraries.

Image-processing software has always been an integral part of structure determination by cryogenic electron microscopy (cryo-EM). Recent advances in hardware and software are recognized as one of the key factors in the so-called cryo-EM resolution revolution. Increasing computational power has opened many possibilities to consider more demanding algorithms, which in turn allow more complex biological problems to be tackled. Moreover, data processing has become more accessible to many experimental groups, with computations that used to last for many days at supercomputing facilities now being performed in hours on personal workstations. All of these advances, together with the rapid expansion of the community, continue to pose challenges and new demands on the software-development side. In this article, the development of emcore and emvis, two basic software libraries for image manipulation and data visualization in cryo-EM, is presented. The main goal is to provide basic functionality organized in modular components that other developers can reuse to implement new algorithms or build graphical applications. An additional aim is to showcase the importance of following established practices in software engineering, with the hope that this could be a first step towards a more standardized way of developing and distributing software in the field.

Adaptive Virtual Organisms: A Compositional Model for Complex Hardware-software Binding*,†

The relation between a structure and the function running on that structure is of central interest in many fields, including computer science, biology (organ vs. function), psychology (body vs. mind), architecture (designs vs. functionality), etc. Our paper addresses this question with reference to computer science recent hardware and software advances, particularly in areas as robotics, AI-hardware, self-adaptive systems, IoT, CPS, etc. At the modelling, conceptual level, our main contribution is the introduction of the concept of (VO), to populate the intermediary level between rigid, slightly reconfigurable, hardware agents and abstract, intelligent, adaptive software agents. A virtual organism has a structure, resembling the hardware capabilities, and it runs low-level functions, implementing the software requirements. The model is compositional in space (allowing the virtual organisms to aggregate into larger organisms) and in time (allowing the virtual organisms to get composed functionalities). Technically, the virtual organisms studied here are in 2D and their structures are described by regular 2D pattens; adding the time dimension, we conclude the VO model is in 3D. By reconfiguration, an organism may change its structure to another structure belonging to the same 2D pattern. We illustrate the VO concept with three increasingly more complex VOs: (1) a tree collector; (2) a feeding cell; and (3) a collection of connected feeding cells. We implemented a simulator for tree collector organisms and the quantitative results show reconfigurable structures are better suited than fixed structures in dynamically changing environments. We briefly show how Agapia - a structured parallel, interactive programming language where dataflow and control flow structures can be freely mixed - may be used for getting quick implementations for VO's simulation.

Superconducting Qubits: Current State of Play

Superconducting qubits are leading candidates in the race to build a quantum computer capable of realizing computations beyond the reach of modern supercomputers. The superconducting qubit modality has been used to demonstrate prototype algorithms in the noisy intermediate-scale quantum (NISQ) technology era, in which non-error-corrected qubits are used to implement quantum simulations and quantum algorithms. With the recent demonstrations of multiple high-fidelity, two-qubit gates as well as operations on logical qubits in extensible superconducting qubit systems, this modality also holds promise for the longer-term goal of building larger-scale error-corrected quantum computers. In this brief review, we discuss several of the recent experimental advances in qubit hardware, gate implementations, readout capabilities, early NISQ algorithm implementations, and quantum error correction using superconducting qubits. Although continued work on many aspects of this technology is certainly necessary, the pace of both conceptual and technical progress in recent years has been impressive, and here we hope to convey the excitement stemming from this progress.

Traversing large graphs on GPUs with unified memory

Due to the limited capacity of GPU memory, the majority of prior work on graph applications on GPUs has been restricted to graphs of modest sizes that fit in memory. Recent hardware and software advances make it possible to address much larger host memory transparently as a part of a feature known as unified virtual memory. While accessing host memory over an interconnect is understandably slower, the problem space has not been sufficiently explored in the context of a challenging workload with low computational intensity and an irregular data access pattern such as graph traversal. We analyse the performance of breadth first search (BFS) for several large graphs in the context of unified memory and identify the key factors that contribute to slowdowns. Next, we propose a lightweight offline graph reordering algorithm, HALO (Harmonic Locality Ordering), that can be used as a pre-processing step for static graphs. HALO yields speedups of 1.5x-1.9x over baseline in subsequent traversals. Our method specifically aims to cover large directed real world graphs in addition to undirected graphs whereas prior methods only account for the latter. Additionally, we demonstrate ties between the locality ordering problem and graph compression and show that prior methods from graph compression such as recursive graph bisection can be suitably adapted to this problem.

Six‐Axis Ground Motion Measurements of Caldera Collapse at Kīlauea Volcano, Hawai'i—More Data, More Puzzles?

AbstractNear‐field recordings of large earthquakes and volcano‐induced events using traditional seismological instrumentation often suffer from unaccounted effects of local tilt and saturation of signals. Recent hardware advances have led to the development of the blueSeis‐3A, a very broadband, highly sensitive rotational motion sensor. We installed this sensor in close proximity to permanently deployed classical instrumentation (i.e., translational seismometer, accelerometer, and tiltmeter) at the Hawaiian Volcano Observatory (USGS). There, we were able to record three ~Mw 5 earthquakes associated with large collapse events during the later phase of the 2018 Kīlauea summit eruption. Located less than 2 km from the origins of these sources, the combined six‐axis translational and rotational measurements revealed clear static rotations around all three coordinate axes. With these six component recordings, we have been able to reconstruct the complete time history of ground motion of a fixed point during an earthquake for the first time.

RiskCog: Unobtrusive Real-Time User Authentication on Mobile Devices in the Wild

Recent hardware advances have led to the development and consumerization of mobile devices, which mainly include smartphones and various wearable devices. To protect the privacy of users, various user authentication mechanisms have been proposed. In particular, biometrics has been widely used for multi-factor authentication. However, biometrics-based authentication mechanisms usually require costly sensors deployed on devices, and rely on explicit user input and Internet connection for performing user authentication. In this article, we propose a system, called RiskCog , which can authenticate the ownership of mobile devices unobtrusively and in a real-time manner by adopting a learning-based approach. Unlike previous studies on user authentication, for cross-platform deployment, maximum user privacy protection, and unobtrusive authentication, RiskCog only relies on those widely available and privacy-insensitive motion sensors to capture the data related to the users’ daily device usage. It requires no users’ explicit input and has no requirement on the users’ motion state or the device placement. RiskCog is also usable in the environment without Internet access by performing offline user identity verification. We conduct comprehensive experiments on smartphones and smartwatches, which show that RiskCog can authenticate device users rapidly and with high accuracy.

A Survey on the Internet of Things (IoT) Forensics: Challenges, Approaches, and Open Issues

Today is the era of the Internet of Things (IoT). The recent advances in hardware and information technology have accelerated the deployment of billions of interconnected, smart and adaptive devices in critical infrastructures like health, transportation, environmental control, and home automation. Transferring data over a network without requiring any kind of human-to-computer or human-to-human interaction, brings reliability and convenience to consumers, but also opens a new world of opportunity for intruders, and introduces a whole set of unique and complicated questions to the field of Digital Forensics. Although IoT data could be a rich source of evidence, forensics professionals cope with diverse problems, starting from the huge variety of IoT devices and non-standard formats, to the multi-tenant cloud infrastructure and the resulting multi-jurisdictional litigations. A further challenge is the end-to-end encryption which represents a trade-off between users' right to privacy and the success of the forensics investigation. Due to its volatile nature, digital evidence has to be acquired and analyzed using validated tools and techniques that ensure the maintenance of the Chain of Custody. Therefore, the purpose of this paper is to identify and discuss the main issues involved in the complex process of IoT-based investigations, particularly all legal, privacy and cloud security challenges. Furthermore, this work provides an overview of the past and current theoretical models in the digital forensics science. Special attention is paid to frameworks that aim to extract data in a privacy-preserving manner or secure the evidence integrity using decentralized blockchain-based solutions. In addition, the present paper addresses the ongoing Forensics-as-a-Service (FaaS) paradigm, as well as some promising cross-cutting data reduction and forensics intelligence techniques. Finally, several other research trends and open issues are presented, with emphasis on the need for proactive Forensics Readiness strategies and generally agreed-upon standards.

Accelerating the Estimation of Metabolic Cost Using Signal Derivatives: Implications for Optimization and Evaluation of Wearable Robots

A chief goal for lower-limb wearable robots (e.g., exoskeletons) is to augment the user's natural motion in a helpful, intuitive way. Accordingly, a common evaluation metric for such systems is the effect of the robotic assistance on the metabolic cost of the wearer. Recent hardware and control advancements have enabled researchers to achieve the challenging task of reducing the metabolic cost of walking below that of walking without an exoskeleton, using autonomous [1], [2], tethered [3], and even passive [4] devices. Although energy expenditure has long been used as an evaluation metric for device efficacy, it has also been proposed as a design specification (i.e., the Augmentation Factor [5]) and, recently, used as a physiological cost function for the realtime control of wearable robotic devices-a strategy known as body-in-the-loop or human-in-the-loop optimization . These algorithms use real-time estimates of metabolic cost to iteratively tune the actuation profile of an assistive device to minimize the wearer?s energetic cost [6]-[10].

Applications of charge-density analysis to the rational design of molecular materials: A mini review on how to engineer optical or magnetic crystals

Low-temperature, high-resolution X-ray diffraction has become a rather established technique to unveil fine details of the electronic structure of molecules and even polymeric units in crystals. The most recent advances in methods, hardware and software have enabled the reconstruction of crystalline charge densities with unprecedented high accuracy. A natural step in the evolution of the field would be to apply these results to the rationalization and prediction of macroscopic response properties of crystals, such as thermal expansion and conductivity, electrical polarization, magnetization, mechanical elasticity, etc. Consequently, charge-density analysis is nowadays not limited to find ever more accurate refinement models, but it has also accepted the challenge of helping to engineer application-oriented molecules and materials. Therefore, in the last years, experimentally estimated charge densities have been of interest not only to the crystallographic community, but also to the broader fields of chemistry and materials science. In particular, detailed insight can be gained on the optics or magnetism of molecular materials when their crystalline macroscopic behavior is correlated with structural and electronic features (as obtained from X-ray diffraction or first-principle quantum-mechanical calculations) of their molecular and sub-molecular building blocks. In this mini review, such relationships are outlined. The emphasis is put on amino acids and coordination polymers as prototypical systems, but the conclusions are general in the sense that they may apply to different classes of molecular systems. This work can also be read as a tutorial review by the newcomers interested in the entanglement between charge-density analysis and crystal engineering.