Architectural Choices Research Articles

Investigating efficient CNN architecture for multiple sclerosis lesion segmentation.

Purpose: The automatic segmentation of multiple sclerosis lesions in magnetic resonance imaging has the potential to reduce radiologists' efforts on a daily time-consuming task and to bring more reproducibility. Almost all new segmentation techniques make use of convolutional neural networks with their own different architecture. Architectural choices are rarely explained. We aimed at presenting the relevance of a U-net-like architecture for our specific task and at building an efficient and simple model. Approach: An experimental study was performed by observing the impact of applying different mutations and deletions to a simple U-net-like architecture. Results: The power of the U-net architecture is explained by the joint benefits of using an encoder-decoder architecture and by linking them with long skip connections. Augmenting the number of convolutional layers and decreasing the number of feature maps allowed us to build an exceptionally light and competitive architecture, the minimally parameterized U-net (MPU-net), with only parameters. Conclusion: The empirical study of the U-net has led to a better understanding of its architecture. It has guided the building of the MPU-net, a model far less parameterized than others (at least by a factor of seven). This neural network achieves a human-level segmentation of multiple sclerosis lesions on fluid-attenuated inversion recovery images only. It shows that this segmentation task does not necessitate overly complicated models to be achieved. This gives the opportunity to build more explainable models that can help such methods to be adopted in a clinical environment.

Использование технологии сверточных нейронных сетей в сегментации объектов изображения

The classification and description of the tasks solved using computer vision technologies are given. The use of neural networks to create systems for selecting objects in an image stream is considered in more detail. It also explains what is meant by training a neural network and discusses in detail the main stages of machine learning. The features of the application of convolutional neural networks for the segmentation of image objects, i.e., the selection of objects in the image, are indicated. The choice of the neural network architecture has been made, which has the property of extracting basic information from the image. The characteristics of the segmentation problem and the basic principles of computer vision are given. Conclusions are given on the possible application of the developed neural network model for solving various applied problems.

No-instruction-set-computer design experience of flexible and efficient architectures for digital communication applications: two case studies on MIMO turbo detection and universal turbo demapping

The emerging flexibility need in designing application-specific processors dedicated for modules of digital receiver imposes a new design metric, which is added to the requirements of efficiency and productivity. In order to cope with the emerging flexibility requirement combined with the best performance efficiency, many application-specific processor design approaches have been proposed and investigated. In general, available design approaches that adopt dynamic scheduling of instructions add an overhead due to the instruction decoding. To minimize this overhead, several approaches have been introduced, which opt static scheduling. In this context, No-Instruction-Set-Computer (NISC) concept has been introduced to design application-specific processors without an instruction set. NISC concept proposes that there is no need to first design and then use an instruction set when the hardware is programmed by its designers rather than its users. NISC designing approach offers a good compromise between flexibility, productivity, and quality for the design of a digital system. In our work, NISC approach is explored through the design of flexible and efficient architectures dedicated for digital communication applications which fulfill the requirements imposed by multiple emergent communication standards. This paper introduces briefly the NISC concept and the corresponding design methodology. Also, it provides an overview of the related design approach. In addition, the relevance of NISC in realizing flexible and efficient implementation in the domain of digital communication is demonstrated through two case studies on MIMO turbo detection and universal turbo demapping. Both designed NISC-based architectures have been compared to state-of-the-art ASIP-based architectures using similar computational resources and supporting same flexibility parameters. The obtained results show that the proposed NISC-based architectures provide a significant improvement in execution performance while having reduced implementation costs. The results also illustrates how the control memory requirements depend on the application and the devised architecture choices. In the detector module, the adopted re-usability of allocated resources imposes separate controlling of each component; hence, additional control signals are implied. Whereas for the demapper module, implemented hardware components are considered to perform specific operations and to deal with the same type of data; hence, the number of control signals can be reduced significantly.

Why vehicle architecture is so important for comfort dynamics? Certain analysis tools of comfort dynamics for vehicle architectures

Ground vehicle architectures are designed to prevent the transmission of vibration, shock, and noise from the ground profile under operating conditions. The choice of a kinematic architecture strongly contributes to the filtering efficiency of the complete system and has to be considered before the efficiency of the suspension components. In order to highlight the filtering contribution of architectures, all flexible (and dissipative) behaviours are removed in studied models: all the components are assumed as rigid bodies. Thus, this paper focuses on the kinematic filtering performed by ground vehicle architectures, especially articulated suspension systems.The proposed methodology for studying the filtering efficiency of ground vehicle architectures is described and illustrated by the study of a single wheel to that of the rocker-bogie system used for the rover Curiosity, passing through the intermediate case of a bicycle. Firstly, an original and robust method is proposed in order to simulate this type of mechanism. For this purpose, the kinematics of the studied system is defined by algebraic equations transformed into a set of ordinary differential equations by using an adaptation of the generalized cross product. Secondly, the nonlinear filtering effect of the kinematic function of the system is studied and discussed. This paper provides a set of nonlinear analysis tools applied to kinematics of a ground vehicle, to characterize and analyse architectural designs.The two main contributions of this work are, first, a kinematic solver method based on generalised cross product, and secondly, dedicated vibration comfort metrics for embedded systems in articulated mechanism (based on nonlinear filtering analysis).

Stand‐Alone Photoelectrochemical Energy Conversions

The photoelectrochemical (PEC) conversions of water and atmospheric CO2 to value‐added fuels, such as H2, CH4, and CH3OH, can provide potential alternative sources for clean and environment‐friendly solar fuels. Several efforts are reported on the designing and developing stand‐alone PEC cells that efficiently produce H2 and O2 from water and minimize atmospheric CO2 via conversion to fuels under only sunlight. However, in reality, high overpotential, poor product‐selectivity, competitive side‐reactions, and self‐reduction of a catalyst limit the performance of the PEC cells, thereby requiring high power inputs. The choice of electrode materials and architectures of PEC cells are very important for designing stand‐alone and durable PEC cells with high solar‐to‐fuels conversion efficiency (EffSTF) and high selectivity. The present review provides a complete account of recent published works on stand‐alone PEC systems with different architectures; development of electrode materials for high EffSTF, selectivity, and stability; and the current challenges. Furthermore, this review describes the future outlook on stand‐alone PEC systems for future production of clean solar fuels and mitigation of atmospheric CO2 levels by utilizing only solar energy.

AutoFrame: A Novel Procedure to Auto-Convert Architectural Massing Models into Structural Simulation Models to Streamline Embodied- and Operational-Carbon Assessment and Daylight Evaluation in Early Design

Sustainability and material efficiency are essential considerations in architecture. However, they are often evaluated late, absent optimization potentials inherent in architectural choices. Easy-to-use computational tools facilitate integration of performance parameters into design decision-making, but because different simulation environments require specific geometric input, simultaneous consideration of multiple constraints is not feasible without significant modeling. This research capitalizes on existing simulation tools and presents a novel procedure, AutoFrame, that converts architectural massing models into structural simulation input models to streamline daylight simulation, and embodied- and operational-carbon assessment during schematic design. Three reference buildings are used to validate the approach and a speculative case study demonstrates how the multi-disciplinary performance feedback guides design decisions while maintaining the flexibility of early design exploration.

Application of neural network in prediction of temperature: a review

The aim of this study was to review different literatures to assess the applicability of artificial neural network in predicting temperature. Temperature prediction as part of weather prediction involves the application of science and technology to predict the state of temperature for a future period in a specific location. Artificial neural network (ANN) has been found to be a promising tool to be used in temperature prediction because it is able to handle complex and nonlinear physical variables of the atmosphere. The use of ANN for prediction of weather elements has shown significant improvements in prediction and accuracy. The performance of the ANN model varies depending on the nature and number of input data used in training the network, the number of neurons in the hidden layer, architecture of a network, transfer function and on the training algorithms. The choice of ANN architecture and the type of data depend on the nature of the problem to be addressed. ANN is therefore found to be a powerful tool in predicting temperature of a specific place, provided input parameters of the model are well chosen.

Automotive Architecture Topologies: Analysis for Safety-Critical Autonomous Vehicle Applications

Safety-critical systems such as Advanced Driving Assistance Systems and Autonomous Vehicles require redundancy to satisfy their safety requirements and to be classified as fail-operational. Introducing redundancy in a system with high data rates and processing requirements also has a great impact on architectural design decisions. The current self-driving vehicle prototypes do not use a standardized system architecture but base their design on existing vehicles and the available components. In this work, we provide a novel analysis framework that allows us to qualitatively and quantitatively evaluate an in-vehicle architecture topology and compare it with others. With this framework, we evaluate different variants of two common topologies: domain and zone-based architectures. Each topology is evaluated in terms of total cost, failure probability, total communication cable length, communication load distribution, and functional load distribution. We introduce redundancy in selected parts of the systems using our automated process provided in the framework, in a safety-oriented design process that enables the ISO26262 Automotive Safety Integrity Level decomposition technique. After every design step, the architecture is re-evaluated. The advantages and disadvantages of the different architecture variants are evaluated to guide the designer towards the choice of correct architecture, with a focus on the introduction of redundancy.

An approach to assessing the technical condition of elements and nodes of transport systems using neural network modeling methods and digital twin technology

The development of intelligent methods and the development of tools for solving research problems of modeling and diagnosing the state of technical systems are relevant areas related to the introduction of digital technologies. Such problems include the problems of preparing correct data arrays for diagnosing and predicting the state of elements and nodes of transport systems. The purpose of the paper is to develop methods for preparing and analyzing data for modeling digital twins on the example of studying the temperature regime of the functioning of a railway car's axle box node, as well as choosing the architecture of neural networks for solving problems of assessing the technical condition of the axle box nodes. The features of the digital twin technology are characterized and the directions of its use for modeling and diagnostics of railway transport systems are considered. As part of the development of an approach to solving the problem of diagnosing the condition of the axle box units of railway cars, analytical dependences and generalized heat transfer equations related to changes in temperature regimes are considered. Examples of the presentation of such data on temperature series, which contain values for deviations from the normal functioning of the elements of the axle box node, are given. The choice of the neural network architecture adapted for solving the problems of estimating and predicting the temperature values of the axle box unit of the car was made. Options for preparing test data for a neural network model have been developed. The results can be used in the problems of creating algorithmic and software for preparing correct arrays of input data for technical diagnostics, in the problems of synthesis and analysis of models of intelligent systems, in various machine learning problems. The considered approach to modeling is aimed at developing methods for assessing and predicting the state of transport infrastructure elements and can be used in the development of intelligent transport systems and for improving the technology of digital twins.

To Reassemble Capitalism: Economic Sociology and its “Political Unconscious.” On “Political Economy after Neoliberalism” by Neil Fligstein and Steven Vogel

The article offers a commentary on the discussion of the article “Political Economy after Neoliberalism” by Fligstein and Vogel, published in the current issue of the Journal of Economic Sociology. The authors draw the attention of Russian-speaking readers to the fact that the work of American researchers not only problematizes the content of political debates in the United States, but also shapes the basic principles of economic and sociological analysis of various economic systems in their connection with regulation policies and public control. Arguments are given in favor of the fact that the article by Fligstein and Vogel is a kind of manifesto of new economic sociology, demonstrating its “political unconscious”-a number of axiomatic assumptions about the functioning of the capitalist political economy, arising from the research perspective of economic sociology and related disciplines. The structure of the argument proposed in the article includes an analysis of several theoretical and empirical directions: a discussion about the varieties of empirical models of capitalism and statements about the political nature of choice of the institutional architecture of economies, the ways of organizing relations between corporations and society, and the role of the state in the economy. The authors note that the so-called neoliberal turn in social and economic policy in recent years was partly based on the purely intellectual principle in mainstream economic theory that opposes states and markets. Studies in the field of economic sociology, history, and comparative political economy demonstrate the fallacy of this statement, offering a conceptual resource for rethinking modern capitalism.

Quantum Circuit Learning to Compute Option Prices and Their Sensitivities

Quantum circuit learning is applied to computing option prices and their sensitivities. The advantage of this method is that a suitable choice of quantum circuit architecture makes it possible to compute the sensitivities analytically by applying parameter-shift rules. We expect our numerical result to pave the way for using quantum machine learning for option pricing.

Choice of architecture for heterogeneous information resources integration systems in natural-technical objects complex modeling

The paper presents a methodology of an architecture type multi-criteria choice for a system of integrating heterogeneous data required for complex modeling of natural-technical objects. Suggested methodology is based on using linguistic scales for architectures quality particular indicators, as well as methods of decisions verbal analysis and the experiment planning theory for processing results of peer review in the final architectures quality assessment.

Analysis on the Study of Integrated Planning Path of Architectural Design and Landscape Design

Throughout the ages, people's pursuit of a unified and harmonious architectural environment has never stopped. When embarking on a construction project, it integrates architectural design and landscape design into integrated planning, fully integrating architectural space, landscape environment and humanistic environment, thereby forming an architectural design that integrates internal and external environments. This article aims to study the integrated planning path of architectural design and landscape design. This article first analyzes the embodiment of the integration of architectural design and landscape design, classifies the buildings in the landscape environment according to different functions and morphological characteristics, and summarizes the relationship between the building and the landscape. This paper studies the integrated planning path of architecture and landscape design from the aspects of integration under the display of landscape commonality, integration of architecture and environment under the interconnection of scenes, etc., and explores effective strategies under specific measures. The survey data of tourists on the ideal choice of architecture and landscape design in Natural Scenic Area A shows that 86.27% of them prefer to preserve the natural scenery, and 50% of them are in favor of abundant supporting facilities. Therefore, natural tourist areas should add more natural elements to retain the characteristics of their natural tourist areas.

A Tale of Two Lines: “The Transylvanian” and “The Imperial”: Mapping Territorial Integration through Railway Architecture

AbstractWhile states undertook railway construction targeting economic and military objectives, this article questions whether and to which extent their symbolic territorial cohesion was also at stake. The hypothesis we aim to verify is that railway buildings acted as recurrent visual signifiers of territorial coherence and had, therefore, the potential of being instrumental as state-building tools. This research explores how an architectural reading of railway networks can inform our understanding of state-building projects and processes. We expect that geographically scoped railway architectural history is capable of cross-fertilizing political and planning history, through a better understanding of empire, state, and regional building discourses. The investigation focuses on the stylistic architectural choices of edifices on two trunk lines in Transylvania, North-West Romania, before World War I, while this territory belonged to the Habsburg then, as of 1867, Austro-Hungarian Empire. The large-scale analysis of railway architecture is discussed in relation to railway-line ownership, political (central, regional, and local) agency, economic development, and architectural Zeitgeist, highlighting state-building and territorial integration patterns. The mapping carried out reveals two successive architectural layers. These denote a shift in the role of railway architecture from an initial liberal phase, before the 1880s, to a bloom phase, prior to World War I. While during the former there was little state control over architectural aspects, during the latter architecture became a foremost representation instrument for the state railway administration. At the same time, the extant railway architecture appears as a palimpsest, a genuinely cross-border, European heritage, documenting the dynamics between imperial, state, regional, and local agencies.

All Your Fake Detector are Belong to Us: Evaluating Adversarial Robustness of Fake-News Detectors Under Black-Box Settings

With the hyperconnectivity and ubiquity of the Internet, the fake news problem now presents a greater threat than ever before. One promising solution for countering this threat is to leverage deep learning (DL)-based text classification methods for fake-news detection. However, since such methods have been shown to be vulnerable to adversarial attacks, the integrity and security of DL-based fake news classifiers are under question. Although many works study text classification under the adversarial threat, to the best of our knowledge, we do not find any work in literature that specifically analyzes the performance of DL-based fake-news detectors under adversarial settings. We bridge this gap by evaluating the performance of fake-news detectors under various configurations under black-box settings. In particular, we investigate the robustness of four different DL architectural choices—multilayer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN) and a recently proposed Hybrid CNN-RNN trained on three different state-of-the-art datasets—under different adversarial attacks (Text Bugger, Text Fooler, PWWS, and Deep Word Bug) implemented using the state-of-the-art NLP attack library, Text-Attack. Additionally, we explore how changing the detector complexity, the input sequence length, and the training loss affect the robustness of the learned model. Our experiments suggest that RNNs are robust as compared to other architectures. Further, we show that increasing the input sequence length generally increases the detector’s robustness. Our evaluations provide key insights to robustify fake-news detectors against adversarial attacks.

A survey on emerging issues in interconnection networks

Nowadays, the interconnection network is considered as an important architectural choice for the future parallel system in many processors owing to the scalable nature of the interconnection networ...

Coronavirus disease 2019: a lesson from history

Viral infections have become an increasingly worsening threat, SARS-Cov-2 is the third coronavirus that emerged during the past twenty years The contagion can also occur by inhalation of aerosolized viruses, capable of floating in the air for long times (in closed spaces) and traveling unknown distances Alesson from history shows that exclusive infectious disease hospitals with multiple building are still a modern architectural choice Individual spaces are needed and this is an important way to reduce the viral load Nowadays, the restrictions should involve not only the interpersonal relationships but also environmental measures © 2020 EDIZIONIMINERVAMEDICA

Fuzzy production network model for quality assessment of an information system based on microservices

This article describes the analysis of the quality of microservice architectures, which are one of the main approaches to the creation and maintenance of modern information systems capable of quickly respond to changes in business demands. The implementation of continuous delivery of software components for dynamic business processes of information systems can be carried out by various sets of microservices, the optimal choice of which is a complex multi-alternative task. The paper presents a review of existing approaches to solving the problem, which showed that the development of models for assessing the quality of microservices of information systems requires further elaboration in terms of accounting for uncertainty in the initial data and modes of operation. The authors have proposed an approach to solving the problem of analyzing the quality of a microservice architecture which is implemented on the basis of a fuzzy production network model. The model allows for comprehensive accounting of various parameters (qualitative and quantitative). The article shows the implementation process of the fuzzy production network that was developed to analyze the functional quality of the microservice architecture for processing customer orders using fuzzy modeling software. The results of the analysis will allow managers and system architects to make an informed choice of the microservice architecture of the information system, as well as use it in their reports when arguing the need for scaling the system and increasing the availability of microservices.

AN EXPERIMENTAL STUDY ON EFFECTIVENESS USAGE OF VARIABLE DESIGN SOLUTIONS BY MAINTENANCE STAGE OF SOFTWARE FOR “SMART-HOME” SYSTEMS

Article considers current state of scientific publications on topic of intellectualization of all main processes for full life cycle development of software systems, approaches to building variable models for “Smart Home” systems. A review of existing approaches to the analysis of quality of construction for such solutions. The Feature-Oriented Development Architecture (FODA) model of smart home system is analyzed, the current problems of smart home systems design are emphasized. The state of publications on topic of quality assessment of systems such as «Smart Home» is analyzed. Then, the proposed approach to experimental assessment of system maintenance is based on the analysis of the total costs required to prepare functionality of system for operation. In the course of experimental calculations, domain model of Smart Home systems was used in two variants, using an integration module and using separate control devices, separately for each of the manufacturers. The proposed approach for evaluating alternative FODA models is based on «maintainability» criterion. The time required to configure the software for a set of individual system modules and to configure them as a single system was used as data. The obtained data, allowing to make a decision on choice of construction architecture of system based on the set domain model, contains a set of devices and the description of functions provided by them. The results of the work should be the basis for the development of a centralized storage for the accumulation, analysis and reuse of the configuration of individual submodules of Smart Home systems and integrated systems that meet certain requirements, resulting in a comprehensive approach to improving the quality of Smart Home systems on the example of the traceability indicator.

Features of the Development of a Neural Network to Automate the Recognition of Steel Defects

The article presents an embodiment of an artificial neural network for recognizing defects in images of steel sheets. Several stages of solving the problem are described: the choice of a development environment, a programming language, and libraries necessary for the implementation; features of data analysis, graphing, histograms, finding dependencies; the selection of a suitable neural network, the choice of neural network architecture, the selection of an algorithm for assessing quality and accuracy; neural network spelling; training and checking accuracy and quality, checking for overfitting (retraining). As development tools, Python language, PyTorch library, Jupyter development environment, convolutional neural network architecture — Unet are proposed. Features of the analysis of input images of steel sheets, features of the implementation of the neural network itself are described. The function of binary cross entropy was chosen as a criterion for assessing accuracy, since it seeks to bring the distribution of the network forecast to the target, fine not only for erroneous predictions, but also for uncertain ones. For additional evaluation, the DICE method was also used. The accuracy of the resulting model is 84 %. The proposed solution can become part of a hardware-software system for automating the recognition of defects on metal sheets.