Successive Refinement Research Articles

Towards Fast and Robust Real Image Denoising With Attentive Neural Network and PID Controller

With the development of deep learning technologies, recent research on real-world noisy image denoising has achieved a considerable improvement in performance. However, a common limitation for existing approaches is the imbalanced trade-off between denoising accuracy and efficiency. To address this problem, we propose a robust and efficient denoiser, called a hierarchical-based PID-attention denoising network (HPDNet), to flexibly deal with the sophisticated noise. The core of our algorithm is the PID-attentive recurrent network (PAR-Net) whose framework mainly consists of the LSTM network and PID controller. PAR-Net inherits the advantages of both the attentive recurrent network and control action, which can encourage more discriminatory feature representations. This learning procedure is implemented within a feedback control system, allowing a faster and more robust means to enhance feature discriminability. Furthermore, by decomposing the noisy image and stacking the PAR-Nets, our PAR-Net can work on a progressively hierarchical framework, and hence obtain multi-scale features and manageable successive refinements. On several widely used datasets, the proposed HPDNet demonstrates high efficiency, while delivering a better perceptually appealing image quality over state-of-the-art image denoising methods.

Nanoparticle enabled high performance high modulus steels

Here we present a novel concept that uses a low volume fraction of nanoparticles to nano-treat Fe-Ti-B melts for a new high performance high modulus steel (HMS). Microstructures of the nano-treated Fe-Ti-B HMS were investigated. The nano-treated HMS consists of fine TiB2 particles and a strengthening phase of unexpected ultrafine Fe2B lamellas. While Fe2B phase is traditionally considered to be detrimental in steel making due to its inherent embrittlement effect, this study shows a successful conversion of the detrimental Fe2B phase into an advantageous one. The nano-treated HMS has significantly higher tensile strength compared to its conventionally made counterpart, 950 MPa vs. 510 MPa, while retaining the high Young's modulus, low density, and ductility. This new concept effectively demonstrates the successful structural refinement in Fe-Ti-B system using nanoparticles, without changing the solidification rate, which eliminates the dilemma between achieving enhanced mechanical properties versus manufacturing for mainstream applications and enables the use of stiff HMS for lightweighting of practical applications. A new mechanism of nanoparticle-induced solute enrichment at nanoscale in alloy melts was also proposed and has implications for new manufacturing routes of metals.

Quantitative assessment of finite-element models for magnetostatic field calculations

We present quantitative means for assessing the numerical accuracy of static magnetic field calculations in finite-element models. Our calculations use the three-dimensional Opera simulation software suite of Dassault Systèmes. Our need to assess the effects of fringe fields requires such a 3D algorithm. While we do discuss and compare our approach to a method of accuracy estimation used in the Opera post-processor, our methods are generally applicable to any model using relaxation techniques in finite-element systems. For purposes of illustration, we present modeling and analysis of two types of quadrupole electromagnets presently in operation in the south arc of the Cornell Electron Storage Ring (CESR). Calculations of field multipole expansion coefficients and numerical deviations from Maxwell’s equations in source-free regions are discussed, with emphasis on the dependence of their accuracy on changes to the finite-element model. Successive refinement steps in the finite-element model for the non-extraction type of CESR south arc quadrupole achieve a reduction in the RMS value of the longitudinal component of the curl vector on the magnet axis by a factor of nearly 70 from 4.38×10−2 T/m to 6.36×10−4 T/m, which is 0.0021% of the field gradient. An accuracy of 2.9% is achieved for a dodecapole coefficient of 2.4×10−4 of that of the quadrupole in the longitudinal integral of the vertical field gradient at a radius of 1 cm.

On Asymptotic Analysis of Energy-Distortion Tradeoff for Low-Delay Transmission Over Gaussian Channels

Asymptotic energy-distortion performance of zero- and low-delay transmission of Gaussian sources over energy-limited Gaussian channels is studied. A lower bound for the leading term in the negative logarithm of the distortion, termed the energy-distortion exponent, is derived through an achievable scheme based on high-resolution quantization coupled with orthogonal signaling. The higher-order term in the negative logarithm of the distortion, termed the energy-distortion dispersion, is optimized while keeping the leading term, the energy-distortion exponent, at its optimal (respectively, the best known) value for the zero-delay (respectively, low-delay) regime. In contrast with the decaying dispersion previously reported in the literature, the proposed coding scheme achieves a constant dispersion. When the scheme is optimized, this constant can be improved with respect to its naïve value, i.e., that achieved by optimizing purely the source coding performance instead of the end-to-end distortion. Lastly, a tradeoff of achievable energy-distortion exponents is derived for broadcast scenarios by extending the point-to-point scheme to include a successive refinement source coder coupled with two rounds of orthogonal signaling. A simple parametric computation algorithm is derived for the tradeoff.

Certifications in the Academic Process Oriented to the Reputational Capital of Brands of Brazilian HEIs


 
 
 
 Considering the analysis of the academic process oriented to the reputational capital of the HEI brands, it arrive three main certifiable objects with impact on the reputation of universities: 'individual teaching advice' - applied to perform the main tasks of the process, the ' self-efficacy '- present in the evaluation of the work plan that guides the student's actions in the process and' disclosure '- which brings academia closer to society and generates a positive internal culture. The objective of this work is to identify how to carry out the certification of these objects. As a research methodology, three quasi-systematic reviews were applied, with successive refinements of articles collected on the CAPES Portal and the results were compiled. It was possible to identify that training followed or monitored by evaluations, and practical tests, also called challenges, are the forms of certification prevalent in this scope. The skills developed from the certifications considered were also listed, with the most mentioned being highlighted, such as the development of social skills, strengthening the commitment to the career and participation in the community of continuous learning; being observed that these favor the reputational capital of university brands.
 
 
 
 

Multi-shot distributed transaction commit

Atomic Commit Problem (ACP) is a single-shot agreement problem similar to consensus, meant to model the properties of transaction commit protocols in fault-prone distributed systems. We argue that ACP is too restrictive to capture the complexities of modern transactional data stores, where commit protocols are integrated with concurrency control, and their executions for different transactions are interdependent. As an alternative, we introduce Transaction Certification Service (TCS), a new formal problem that captures safety guarantees of multi-shot transaction commit protocols with integrated concurrency control. TCS is parameterized by a certification function that can be instantiated to support common isolation levels, such as serializability and snapshot isolation. We then derive a provably correct crash-resilient protocol for implementing TCS through successive refinement. Our protocol achieves a better time complexity than mainstream approaches that layer two-phase commit on top of Paxos-style replication.

Marine-Derived Natural Products as ATP-Competitive mTOR Kinase Inhibitors for Cancer Therapeutics.

The mammalian target of rapamycin (mTOR) is a serine/threonine kinase portraying a quintessential role in cellular proliferation and survival. Aberrations in the mTOR signaling pathway have been reported in numerous cancers including thyroid, lung, gastric and ovarian cancer, thus making it a therapeutic target. To attain this objective, an in silico investigation was designed, employing a pharmacophore modeling approach. A structure-based pharmacophore (SBP) model exploiting the key features of a selective mTOR inhibitor, Torkinib directed at the ATP-binding pocket was generated. A Marine Natural Products (MNP) library was screened using SBP model as a query. The retrieved compounds after consequent drug-likeness filtration were subjected to molecular docking with mTOR, thus revealing four MNPs with better scores than Torkinib. Successive refinement via molecular dynamics simulations demonstrated that the hits formed crucial interactions with key residues of the pocket. Furthermore, the four identified hits exhibited good binding free energy scores through MM-PBSA calculations and the subsequent in silico toxicity assessments displayed three hits deemed essentially non-carcinogenic and non-mutagenic. The hits presented in this investigation could act as potent ATP-competitive mTOR inhibitors, representing a platform for the future discovery of drugs from marine natural origin.

A Comprehensive Study of Ca9Tb(PO4)7 and Ca9Ho(PO4)7 Doped β-Tricalcium Phosphates: Ab initio Crystal Structure Solution, Rietveld Analysis, and Dielectric Properties

The scientific interest toward the structural, luminescence, and dielectric properties of tricalcium phosphate (TCP) materials, in particular, the doped TCP-based compounds, is mainly due to their biocompatibility and bioactivity behavior and the increasing number of their applications. In the literature, structural investigations of a doped β-TCP compound are usually carried out by powder diffraction data under the assumption that it has a centrosymmetric whitlockite structure. Recent studies have shown that this assumption is not always fulfilled. For this reason, it is necessary to use methods of nonlinear and dielectric spectroscopy to prove the true symmetry of whitlockite-like phases. This work provides a detailed and comprehensive study on the structural, dielectric, and nonlinear properties of the polycrystalline Tb- and Ho-doped tricalcium phosphates, chemically Ca9Tb(PO4)7 and Ca9Ho(PO4)7. Their crystal structure was explored using powder X-ray diffraction, adopting an ab initio approach for the structure solution followed by the Rietveld refinement. For Ca9Tb(PO4)7, the qualitative analysis of the experimental powder X-ray diffraction pattern revealed that the sample was not a single phase. Nevertheless, its crystal structure has been determined by Direct Methods. The presence of the dopant cation was omitted in the solution process by Direct Methods and considered in the successive Rietveld refinement step. The analysis of the occupancies has clarified the Tb3+ cation distribution in its preferential sites. The same pathway has been followed for the Ca9Ho(PO4)7 sample, which corresponded to a single-phase structure. The dielectric properties of the two doped compounds (Tb3+ and Ho3+) were also examined, and reversible ferroelectric-paraelectric phase transitions were found at 862 and 872 K, respectively.

Generation of synchronized configuration space trajectories with workspace path constraints for an ensemble of robots

We present an approach to generate path-constrained synchronous motion for the coupled ensemble of robots. In this article, we refer to serial-link manipulators and mobile bases as robots. We assume that the relative motion constraints among the objects in the environment are given. We represent the motion constraints as path constraints and pose the problem of path-constrained synchronous trajectory generation as a non-linear optimization problem. Our approach generates configuration space trajectories for the robots to manipulate the objects such that the given motion constraints among the objects are satisfied. We present a method that formulates the problem as a discrete parameter optimization problem and solves it using successive constraint refinement techniques. The method adaptively selects the parametric representation of the configuration variables for a given scenario. It also generates an approximate solution as the starting point for the successive constraint refinement stages to reduce the computation time. We discuss in detail why successive constraint refinement strategies are useful for solving this class of problems. We demonstrate the effectiveness of the proposed method on challenging test cases in simulation and physical environments with high-degree-of-freedom robotic systems.

Exploration of Principal Component Analysis: Deriving Principal Component Analysis Visually Using Spectra.

Spectroscopy rapidly captures a large amount of data that is not directly interpretable. Principal component analysis is widely used to simplify complex spectral datasets into comprehensible information by identifying recurring patterns in the data with minimal loss of information. The linear algebra underpinning principal component analysis is not well understood by many applied analytical scientists and spectroscopists who use principal component analysis. The meaning of features identified through principal component analysis is often unclear. This manuscript traces the journey of the spectra themselves through the operations behind principal component analysis, with each step illustrated by simulated spectra. Principal component analysis relies solely on the information within the spectra, consequently the mathematical model is dependent on the nature of the data itself. The direct links between model and spectra allow concrete spectroscopic explanation of principal component analysis , such as the scores representing "concentration" or "weights". The principal components (loadings) are by definition hidden, repeated and uncorrelated spectral shapes that linearly combine to generate the observed spectra. They can be visualized as subtraction spectra between extreme differences within the dataset. Each PC is shown to be a successive refinement of the estimated spectra, improving the fit between PC reconstructed data and the original data. Understanding the data-led development of a principal component analysis model shows how to interpret application specific chemical meaning of the principal component analysis loadings and how to analyze scores. A critical benefit of principal component analysis is its simplicity and the succinctness of its description of a dataset, making it powerful and flexible.

Severity grading system for acute allergic reactions: A multidisciplinary Delphi study

There is no widely adopted severity grading system for acute allergic reactions, including anaphylactic and nonanaphylactic reactions, thus limiting the ability to optimize and standardize management practices and advance research. The aim of this study was to develop a severity grading system for acute allergic reactions for use in clinical care and research. From May to September 2020, we convened a 21-member multidisciplinary panel of allergy and emergency care experts; 9 members formed a writing group to critically appraise and assess the strengths and limitations of prior severity grading systems and develop the structure and content for an optimal severity grading system. The entire study panel then revised the grading system and sought consensus by utilizing Delphi methodology. The writing group recommended that an optimal grading system encompass the severity of acute allergic reactions on a continuum from mild allergic reactions to anaphylactic shock. Additionally, the severity grading system must be able to discriminate between clinically important differences in reaction severity to be relevant in research while also being intuitive and straightforward to apply in clinical care. Consensus was reached for all elements of the proposed severity grading system. We developed a consensus severity grading system for acute allergic reactions, including anaphylactic and nonanaphylactic reactions. Successful international validation, refinement, dissemination, and application of the grading system will improve communication among providers and patients about the severity of allergic reactions and will help advance future research.

PDWN: Pyramid Deformable Warping Network for Video Interpolation

Video interpolation aims to generate a non-existent intermediate frame given the past and future frames. Many state-of-the-art methods achieve promising results by estimating the optical flow between the known frames and then generating the backward flows between the middle frame and the known frames. However, these methods usually suffer from the inaccuracy of estimated optical flows and require additional models or information to compensate for flow estimation errors. Following the recent development in using deformable convolution (DConv) for video interpolation, we propose a light but effective model, called Pyramid Deformable Warping Network (PDWN). PDWN uses a pyramid structure to generate DConv offsets of the unknown middle frame with respect to the known frames through coarse-to-fine successive refinements. Cost volumes between warped features are calculated at every pyramid level to help the offset inference. At the finest scale, the two warped frames are adaptively blended to generate the middle frame. Lastly, a context enhancement network further enhances the contextual detail of the final output. Ablation studies demonstrate the effectiveness of the coarse-to-fine offset refinement, cost volumes, and DConv. Our method achieves better or on-par accuracy compared to state-of-the-art models on multiple datasets while the number of model parameters and the inference time are substantially less than previous models. Moreover, we present an extension of the proposed framework to use four input frames, which can achieve significant improvement over using only two input frames, with only a slight increase in the model size and inference time.

Mathematical model and algorithm for solving the problem of planning the operation of multiphase systems with heterogeneous resources and time limits

The object of this research is the modern service and production systems, the specific functioning of which lies in a set of sequential and parallel operations with a random duration. A fundamental peculiarity of such systems is the stochastic nature of the duration of a single operation, which depends not only on the external random factors, but also on the choice of resources, and namely on the operator. This substantiates the parallel solution of the task on making a schedule of mutually dependent operations and the task on assigning the operators. In the conditions of resource and time limits, this task is NP difficult and requires the development of algorithms for developing the solution that is close to optimal in the limited time. For the development of mathematical and algorithmic software to solve this task, the author used the critical path method and PERT method, incident wave method, and methods for solving the assignment tasks. As a result, the author acquired a mathematical model that considers the stochastic nature of the duration of a single operation, which depends not only on random factors, but also on the operators. Based on such model, is formulated the optimization task that allows finding the launch time and the corresponding operators to gain the most profit. Based on the analysis of existing approaches and the specificity of the task at hand, the author proposes the algorithm for solving the task founded on successive refinement of the time characteristics of operations.

Weighted-Sum-Rate Maximization for an Reconfigurable Intelligent Surface Aided Vehicular Network

Recently, reconfigurable intelligent surfaces (RISs) have been identified as one potential solution to avoid performance degradation of using millimeter wave (mmWave) frequencies in vehicular communications. In this paper, we investigate the use of an RIS in a mmWave vehicular communication network. The problem of weighted sum-rate maximization in the uplink is considered, where an RIS is used to assist the communication. We focus on both single-user and multi-user cases. Single user case is solved using successive refinement algorithm, where two phase-optimization schemes that help reducing the channel estimation overhead are considered. In multi-user case, fractional programming technique is used to reformulate the original problem into a more convenient form, and an algorithm based on alternating optimization is proposed. The validity of the proposed methods is confirmed by numerical simulations. A significant performance increase is seen when utilizing an RIS in both cases. Considered phase optimization schemes are shown to result in a significant reduction in channel estimation at a cost of small performance degradation compared to the full channel state information beamforming scenario. We perform simulations to investigate the effects of mobility, and the results demonstrate the ability of an RIS to mitigate the effects of mobility to some extent. Furthermore, to get practical insights into vehicular communications aided by an RIS, we use a commercial ray-tracing tool to evaluate the performance.

Ligand-Binding-Site Refinement to Generate Reliable Holo Protein Structure Conformations from Apo Structures.

The first important step in a structure-based virtual screening is the judicious selection of a receptor protein. In cases where the holo protein receptor structure is unavailable, significant reduction in virtual screening performance has been reported. In this work, we present a robust method to generate reliable holo protein structure conformations from apo structures using molecular dynamics (MD) simulation with restraints derived from holo structure binding-site templates. We perform benchmark tests on two different datasets: 40 structures from a directory of useful decoy-enhanced (DUD-E) and 84 structures from the Gunasekaran dataset. Our results show successful refinement of apo binding-site structures toward holo conformations in 82% of the test cases. In addition, virtual screening performance of 40 DUD-E structures is significantly improved using our MD-refined structures as receptors with an average enrichment factor (EF), an EF1% value of 6.2 compared to apo structures with 3.5. Docking of native ligands to the refined structures shows an average ligand root mean square deviation (RMSD) of 1.97 Å (DUD-E dataset and Gunasekaran dataset) relative to ligands in the holo crystal structures, which is comparable to the self-docking (i.e., docking of the native ligand back to its crystal structure receptor) average, 1.34 Å (DUD-E dataset) and 1.36 Å (Gunasekaran dataset). On the other hand, docking to the apo structures yields an average ligand RMSD of 3.65 Å (DUD-E) and 2.90 Å (Gunasekaran). These results indicate that our method is robust and can be useful to improve virtual screening performance of apo structures.

Crystal Structure and Non-Hydrostatic Stress-Induced Phase Transition of Urotropine Under High Pressure.

High-pressure behavior of hexamethylenetetramine (urotropine) was studied in situ using angle-dispersive single-crystal synchrotron X-ray diffraction (XRD) and Fourier-transform infrared absorption (FTIR) spectroscopy. Experiments were conducted in various pressure-transmitting media to study the effect of deviatoric stress on phase transformations. Up to 4 GPa significant damping of molecular librations and atomic thermal motion was observed. A first-order phase transition to a tetragonal structure was observed with an onset at approximately 12.5 GPa and characterized by sluggish kinetics and considerable hysteresis upon decompression. However, it occurs only in non-hydrostatic conditions, induced by deviatoric or uniaxial stress in the sample. This behavior finds analogies in similar cubic crystals built of highly symmetric cage-like molecules and may be considered a common feature of such systems. DFT computations were performed to model urotropine equation of state and pressure dependence of vibrational modes. The first successful Hirshfeld atom refinements carried out for high-pressure diffraction data are reported. The refinements yielded more realistic C-H bond lengths than the independent atom model even though the high-pressure diffraction data are incomplete.

Successive Refinement of Privacy

This work examines a novel question: how much randomness is needed to achieve local differential privacy (LDP)? A motivating scenario is providing multiple levels of privacy to multiple analysts, either for distribution or for heavy hitter estimation, using the same (randomized) output. We call this setting successive refinement of privacy , as it provides hierarchical access to the raw data with different privacy levels. For example, the same randomized output could enable one analyst to reconstruct the input, while another can only estimate the distribution subject to LDP requirements. This extends the classical Shannon (wiretap) security setting to local differential privacy. We provide (order-wise) tight characterizations of privacy-utility-randomness trade-offs in several cases for distribution estimation, including the standard LDP setting under a randomness constraint. We also provide a non-trivial privacy mechanism for multi-level privacy. Furthermore, we show that we cannot reuse random keys over time while preserving privacy of each user.

Space Emerges from What We Know-Spatial Categorisations Induced by Information Constraints.

Seeking goals carried out by agents with a level of competency requires an “understanding” of the structure of their world. While abstract formal descriptions of a world structure in terms of geometric axioms can be formulated in principle, it is not likely that this is the representation that is actually employed by biological organisms or that should be used by biologically plausible models. Instead, we operate by the assumption that biological organisms are constrained in their information processing capacities, which in the past has led to a number of insightful hypotheses and models for biologically plausible behaviour generation. Here we use this approach to study various types of spatial categorizations that emerge through such informational constraints imposed on embodied agents. We will see that geometrically-rich spatial representations emerge when agents employ a trade-off between the minimisation of the Shannon information used to describe locations within the environment and the reduction of the location error generated by the resulting approximate spatial description. In addition, agents do not always need to construct these representations from the ground up, but they can obtain them by refining less precise spatial descriptions constructed previously. Importantly, we find that these can be optimal at both steps of refinement, as guaranteed by the successive refinement principle from information theory. Finally, clusters induced by these spatial representations via the information bottleneck method are able to reflect the environment’s topology without relying on an explicit geometric description of the environment’s structure. Our findings suggest that the fundamental geometric notions possessed by natural agents do not need to be part of their a priori knowledge but could emerge as a byproduct of the pressure to process information parsimoniously.

Minimizing the average achievable distortion using multi-layer coding approach in two-hop networks

Minimizing the average achievable distortion (AAD) of a Gaussian source at the destination of a two-hop block fading relay channel is studied in this paper. The communication is carried out through the use of a Decode and Forward (DF) relay with no source-destination direct links. The associated receivers of both hops are assumed to be aware of the corresponding channel state information (CSI), while the transmitters are unaware of their corresponding CSI. The current paper explores the effectiveness of incorporating the successive refinement source coding together with multi-layer channel coding in minimizing the AAD. In this regard, the closed form and optimal power allocation policy across code layers of the second hop is derived, and using a proper curve fitting approach, a close-to-optimal power allocation policy associated with the first hop is devised. It is numerically shown that the DF strategy closely follows the Amplify and Forward (AF) relaying, while there is a sizable gap between the AAD of multi-layer coding and single-layer source coding.

Adaptive greedy algorithms based on parameter‐domain decomposition and reconstruction for the reduced basis method

AbstractThe reduced basis method (RBM) empowers repeated and rapid evaluation of parametrized partial differential equations through an offline–online decomposition, a.k.a. a learning‐execution process. A key feature of the method is a greedy algorithm repeatedly scanning the training set, a fine discretization of the parameter domain, to identify the next dimension of the parameter‐induced solution manifold along which we expand the surrogate solution space. Although successfully applied to problems with fairly high parametric dimensions, the challenge is that this scanning cost dominates the offline cost due to it being proportional to the cardinality of the training set which is exponential with respect to the parameter dimension. In this work, we review three recent attempts in effectively delaying this curse of dimensionality, and propose two new hybrid strategies through successive refinement and multilevel maximization of the error estimate over the training set. All five offline‐enhanced methods and the original greedy algorithm are tested and compared on two types of problems: the thermal block problem and the geometrically parameterized Helmholtz problem.