Search NASA⌕ Search

SEARCH · Search NASA

Results for “Regularization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 343 records · Page 19

Regularizing INR with Diffusion Prior for Self-Supervised 3D Reconstruction OF Neutron Computed Tomography Data

Recently, generative diffusion priors have made huge strides as inverse problem solvers, including the ability to be adapted for inference on out-of-distribution data. Concurrently, implicit neural representations (INRs) have emerged as fast and lightweight inverse imaging solvers that are amenable to hybrid approaches that combine learned priors with traditional inverse problem formulations. In this paper, we present a diffusive computed tomography (CT) inversion framework for regularizing INRs called Diffusive INR (DINR), designed to enable high-quality reconstruction from sparse-view neutron CT. Pretrained purely on synthetic data, DINR is evaluated on simulated and experimentally obtained observations of concrete microstructures, where traditional reconstruction methods suffer substantial degradation when the number of views is reduced. Our approach delivers superior performance, reduces reconstruction artifacts, and achieves gains in PSNR and SSIM, enabling accurate micro-structural characterization even under extreme data limitations compared to state-of-the-art sparse-view reconstruction techniques.

Hossain, Maliha [ORNL]↗

DENSECL: Haze Mitigation Using Dense Blocks and Contrastive Loss Regularization

Haze, which occurs as a result of the scattering of light in the atmosphere by small particles, diminishes the visibility of scene objects, inflicting important image applications such as object detection. To address the problem, this paper introduces a new physics-based end-to-end deep learning approach to haze mitigation in outdoor scenes, including those in airborne images. The proposed model named DenseCL is designed with dense blocks and adopts a contrastive loss function as an additional regularization. The model also maintains the cycle consistency by remapping the dehazed outputs into a hazy image using the physics-based light scattering function. DenseCL has been trained with publicly available outdoor images and demonstrates outstanding performance on outdoor, indoor, and remotely sensed nonhomogeneous haze satellite images.

Mitra, Somosmita↗

REV-INR: Regularized Evidential Implicit Neural Representation for Uncertainty-Aware Volume Visualization

Applications of Implicit Neural Representations (INRs) have emerged as a promising deep learning approach for compactly representing large volumetric datasets. These models can act as surrogates for volume data, enabling efficient storage and on-demand reconstruction via model predictions. However, conventional deterministic INRs only provide value predictions without insights into the model’s prediction uncertainty or the impact of inherent noisiness in the data. This limitation can lead to unreliable data interpretation and visualization due to prediction inaccuracies in the reconstructed volume. Identifying erroneous results extracted from model-predicted data may be infeasible, as raw data may be unavailable due to its large size. To address this challenge, we introduce REV-INR, Regularized Evidential Implicit Neural Representation, which learns to predict data values accurately along with the associated coordinate-level data uncertainty and model uncertainty using only a single forward pass of the trained REV-INR during inference. By comprehensively comparing and contrasting REV-INR with existing well-established deep uncertainty estimation methods, we show that REV-INR achieves the best volume reconstruction quality with robust data (aleatoric) and model (epistemic) uncertainty estimates using the fastest inference time. Consequently, we demonstrate that REV-INR facilitates assessment of the reliability and trustworthiness of the extracted isosurfaces and volume visualization results, enabling analyses to be solely driven by model-predicted data.

Saklani, Shanu [Indian Institute of Technology, Ka↗

Enhancing Solar Power Forecasting with Regularized Constrained Quantile Regression Averaging and Bootstrapping Techniques

Probabilistic solar power forecasting (SPF) plays an essential role in optimizing power-grid operations by quantifying the forecast uncertainty. To improve the accuracy and robustness of probabilistic SPF, this paper introduces the regularized constrained quantile regression averaging (rCQRA) method to combine outputs from multiple PSPF models. In addition, a bootstrapping method was used to quantify model uncertainty, providing insights into the reliability and significance of each ensemble component. To evaluate its efficacy, the proposed rCQRA method is used to integrate four PSPF methods. The resulting SPF models are trained and validated using a real-world six-year dataset from a rooftop solar plant in the USA. The performance of the proposed rCQRA method is evaluated and compared with two benchmark methods under three categories of weather conditions. It is shown that the rCQRA method has superior performance in its forecast reliability, sharpness, and accuracy.

Ensemble learning, probabilistic solar power forec↗

O3BNN-R: An Out-Of-Order Architecture for High-Performance and Regularized BNN Inference

Binarized Neural Networks (BNN) have drawn tremendous attention due to significantly reduced computational complexity and memory demand. They have especially shown great potential in cost- and power-restricted domains, such as IoT and smart edge-devices, where reaching a certain accuracy bar is often sufficient, and real-time is highly desired.In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, an out-of-order (OoO) architecture – O3BNN-R, can curtail edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism(ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be difficult to exploit. In order to increase the pruning opportunities, we also optimize the training process by adding 2 regularization items in the loss function (1) for pooling pruning and (2) for threshold pruning. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10.

Geng, Tong↗

Hydrodynamic Analysis of a Suspended Cylinder Under Regular Wave Loading Based on Computational Fluid Dynamics

An investigation into the computation of hydrodynamic loads on a suspended cylinder in regular waves is presented. The primary goal was to perform a three-way validation of the loads between experimental measurements and simulations from two computational methods. Experimental measurements of the longitudinal in-line force on a cylinder suspended at a fixed position were available from the Offshore Code Comparison Collaboration, Continued, with Correlation (OC5) project, Phase Ia. These measurements were compared to computational fluid dynamics (CFD) simulations based on the solution of Reynolds-averaged Navier-Stokes (RANS) equations, as implemented in STAR-CCM+. The study encompassed a sensitivity analysis of the loads computed in STAR-CCM+ based on wave modeling, boundary conditions, turbulence modeling, and spatial and temporal discretization. The analysis was supplemented by results generated with the offshore wind turbine engineering software OpenFAST, based on a hybrid combination of second-order potential flow and viscous drag from Morison's equation. The focus of the investigation was on the assessment of the accuracy of the computation of first- and higher-order hydrodynamic loads. Substantial differences were observed in the numerical prediction of the second and third harmonic force contribution. Local flow field analysis with CFD was applied to study the physics of wave run-up and diffraction dynamics to identify the causes.

16 TIDAL AND WAVE POWER↗

Inverse planning for IMRT with nonuniform beam profiles using total-variation regularization (TVR)

Purpose: Radiation therapy with high dose rate and flattening filter-free (FFF) beams has the potential advantage of greatly reduced treatment time and out-of-field dose. Current inverse planning algorithms are, however, not customized for beams with nonuniform incident profiles and the resultant IMRT plans are often inefficient in delivery. The authors propose a total-variation regularization (TVR)-based formalism by taking the inherent shapes of incident beam profiles into account. Methods: A novel TVR-based inverse planning formalism is established for IMRT with nonuniform beam profiles. The authors introduce a TVR term into the objective function, which encourages piecewise constant fluence in the nonuniform FFF fluence domain. The proposed algorithm is applied to lung and prostate and head and neck cases and its performance is evaluated by comparing the resulting plans to those obtained using a conventional beamlet-based optimization (BBO). Results: For the prostate case, the authors' algorithm produces acceptable dose distributions with only 21 segments, while the conventional BBO requires 114 segments. For the lung case and the head and neck case, the proposed method generates similar coverage of target volume and sparing of the organs-at-risk as compared to BBO, but with a markedly reduced segment number. Conclusions: TVR-based optimization in nonflat beam domain provides an effective way to maximally leverage the technical capacity of radiation therapy with FFF fields. The technique can generate effective IMRT plans with improved dose delivery efficiency without significant deterioration of the dose distribution.

60 APPLIED LIFE SCIENCES↗

SU-F-303-02: Achieving 4D MRI in Regular Breathing Cycle with Extended Acquisition Time of Dynamic MR Images

Purpose: The purpose of this study is to develop a technique to obtain four-dimensional (4D) magnetic resonance (MR) images that are more representative of a patient’s typical breathing cycle by utilizing an extended acquisition time while minimizing the image artifacts. Methods: The 4D MR data were acquired with the balanced steady state free precession in two-dimensional sagittal plane of view. Each slice was acquired repeatedly for about 15 s, thereby obtaining multiple images at each of the 10 phases in the respiratory cycle. This improves the probability that at least one of the images were acquired at the desired phase during a regular breathing cycle. To create optimal 4D MR images, an iterative approach was used to identify the set of images that yielded the highest slice-to-slice similarity. To assess the effectiveness of the approach, the data set was truncated into periods of 7 s (50 time points), 11 s (75 time points) and the full 15 s (100 time points). The 4D MR images were then sorted with data of the three different acquisition periods for comparison. Results: In general, the 4D MR images sorted using data from longer acquisition periods showed less mismatched artifacts. In addition, the normalized cross correlation (NCC) between slices of a 4D volume increases with increased acquisition period. The average NCC was 0.791 from the 7 s period, 0.794 from the 11 s period and 0.796 from the 15 s period. Conclusion: Our preliminary study showed that extending the acquisition time with the proposed sorting technique can improve image quality and reduce artifact presence in the 4D MR images. Data acquisition over two breathing cycles is a good trade-off between artifact reduction and scan time. This research was partially funded by the the Center for Radiation Oncology Research from UT MD Anderson Cancer Center.

60 APPLIED LIFE SCIENCES↗

Effective regularized algebraic reconstruction technique for computed tomography

A new fast version of the reconstruction algorithm for computed tomography based on the simultaneous algebraic reconstruction technique (SART) is proposed. The algorithm iteration is asymptotically accelerated using the fast Hough transform from O(n{sup 3}) to O(n{sup 2}logn). Similarly to the algebraic reconstruction technique (RegART), which was proposed by us previously, the regularization operator is applied after each iteration. A bilateral filter plays the role of this operator. The algorithm behavior is investigated using the model experiment.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Analysis of regularizing properties of nonlinear electrodynamics in the Einstein-Born-Infeld theory

New regularizing manifestations of the Einstein-Born-Infeld theory for a massive charged force center in the space-time are considered. The properties of isotropic geodesics in this space are analyzed. It is shown that the charge may exceed the maximum possible charge in the Reissner-Nordstroem solution; the possibility of eliminating one of the metric horizons is also noted.

79 ASTRONOMY AND ASTROPHYSICS↗

Relation between two-point Green’s functions of N=1 SQED with N{sub f} flavors, regularized by higher derivatives, in the three-loop approximation

We verify the identity which relates the two-point Green’s functions of N=1 SQED with N{sub f} flavors, regularized by higher derivatives, by explicit calculations in the three-loop approximation. This identity explains why, in the limit of the vanishing external momentum, the two-point Green’s function of the gauge superfield is given by integrals of double total derivatives in the momentum space. It also makes it possible to exactly derive the NSVZ β-function in all loops if the renormalization group functions are defined in terms of the bare coupling constant. In order to verify the considered identity, we use it to construct integrals giving the three-loop β-function starting from the two-point Green’s functions of matter superfields in the two-loop approximation. Then we demonstrate that the results for these integrals coincide with the sums of the corresponding three-loop supergraphs.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Regular and irregular dynamics of spin-polarized wavepackets in a mesoscopic quantum dot at the edge of topological insulator

The dynamics of Dirac–Weyl spin-polarized wavepackets driven by a periodic electric field is considered for the electrons in a mesoscopic quantum dot formed at the edge of the two-dimensional HgTe/CdTe topological insulator with Dirac–Weyl massless energy spectra, where the motion of carriers is less sensitive to disorder and impurity potentials. It is observed that the interplay of strongly coupled spin and charge degrees of freedom creates the regimes of irregular dynamics in both coordinate and spin channels. The border between the regular and irregular regimes determined by the strength and frequency of the driving field is found analytically within the quasiclassical approach by means of the Ince–Strutt diagram for the Mathieu equation, and is supported by full quantum-mechanical simulations of the driven dynamics. The investigation of quasienergy spectrum by Floquet approach reveals the presence of non-Poissonian level statistics, which indicates the possibility of chaotic quantum dynamics and corresponds to the areas of parameters for irregular regimes within the quasiclassical approach. We find that the influence of weak disorder leads to partial suppression of the dynamical chaos. Our findings are of interest both for progress in the fundamental field of quantum chaotic dynamics and for further experimental and technological applications of spindependent phenomena in nanostructures based on topological insulators.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Tidal Effects in Some Regular Black Holes

This paper is aimed to study the tidal forces produced by a class of regular black holes. We consider the radial infall of test particle and find radial as well as angular components of tidal forces by taking geodesic deviation equations. We also compute geodesic deviation vector by solving geodesic deviation equation numerically. It is concluded that a particle undergos either compression or stretching in radial or angular direction due to tidal forces.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

2 + 1 Flavors QCD Equation of State in NJL Model with Proper Time Regularization

In this paper, we firstly combine the 2 + 1 flavors Nambu-Jona-Lasinio (NJL) model with propertime regularization (PTR) to study the equation of state (EOS) and fix the corresponding parameters in 2 + 1 flavors system. Based on this model, we calculate chiral susceptibility and quark number susceptibility and we find a crossover with chemical potential from 220 to 400 MeV. Then we consider the chemical equilibrium and electric charge neutrality conditions in electroweak reactions, which give some constrains of chemical potentials of different quark flavors. At last, we find that the 2 + 1 flavors system is more stable than 2 flavors system after a comparison and discussion of these two cases.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

On the possible dynamical realization of the Pauli–Villars regularization

The problem of free-particle scattering on virtual wormholes is considered. It is shown that, for all types of relativistic fields, this scattering leads to the appearance of additional very heavy particles, which play the role of auxiliary fields in the invariant scheme of Pauli–Villars regularization. A nonlinear correction that describes the back reaction of particles to the vacuum distribution of virtual wormholes is also obtained.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Polarimetry of Regular and Stochastic Signals in Magnetooptics

A brief review is presented of application of the high-sensitive laser polarimetry for detecting small changes—either regular, or stochastic—of the medium magnetization. Principles and results of application of the spectroscopy of spin fluctuations to atomic and semiconductor systems are described. New possibilities of studying the radiospectroscopic and optical properties of spin systems using the methods of noise spectroscopy are considered. Prospects for expanding the circle of objects of the spin noise spectroscopy at the expense of impurity dielectrics are discussed.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Regular Microstructures Based on Polyethylene Terephthalate Films

A method for forming regular dielectric and metallic microstructures based on polyethylene terephthalate films irradiated with a synchrotron radiation source through a mask with subsequent etching. To study their geometric parameters, scanning electron microscopy was used.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Moment Representation of Regularized Lattice Boltzmann Methods on NVIDIA and AMD GPUs

The lattice Boltzmann method is a highly scalable Navier-Stokes solver that has been applied to flow problems in a wide array of domains. However, the method is bandwidth-bound on modern GPU accelerators and has a large memory footprint. In this paper, we present new 2D and 3D GPU implementations of two different regularized lattice Boltzmann methods, which are not only able to achieve an acceleration of ∼ 1.4 × w.r.t. reference lattice Boltzmann implementations but also reduce the memory requirements by up to 35% and 47% in 2D and 3D simulations respectively. These new approaches are evaluated on NVIDIA and AMD GPU architectures.

Valero Lara, Pedro↗