Search

We propose a method to obtain diagonal mass matrices for NURBS-based approximation spaces by a “dual lumping” method. The use of lumped mass matrices is of great importance in elastodynamics problems, as they can be employed in explicit time integration schemes which do not require the solution of a linear system. In finite elements, several well-established methods, such as row-sum, diagonal scaling, or nodal quadrature methods have been used to obtain lumped mass matrices for different applications. However, for higher-order and higher continuity approximation spaces such as those derived from NURBS, these approaches have only limited (second-order) accuracy. In this work, we derive a dual basis which has optimal approximation and dispersion properties, while maintaining local support. The dual space has discontinuities at the element boundaries (knots) and it is used to provide the test functions in the context of a Petrov–Galerkin method. This results in a general framework for the study of lumped mass matrices which can be employed in explicit time integration schemes with high-order accuracy. Numerical experiments are presented to demonstrate the applicability of the method to problems with smooth solutions as well as to wave propagation problems with reduced regularity.

We propose a computational methodology to perform inverse design of quantum spin hall effect (QSHE)-based phononic topological insulators. We first obtain two-fold degeneracy, or a Dirac cone, in the band structure using a level set-based topology optimization approach. Subsequently, four-fold degeneracy, or a double Dirac cone, is obtained by using zone folding, after which breaking of translational symmetry, which mimics the effect of strong spin-orbit coupling and which breaks the four-fold degeneracy resulting in a bandgap, is applied. We use the approach to perform inverse design of hexagonal unit cells of C6 and C3 symmetry. The numerical examples show that a topological domain wall with two variations of the designed metamaterials exhibit topologically protected interfacial wave propagation, and also demonstrate that larger topologically-protected bandgaps may be obtained with unit cells based on C3 symmetry.

Vehicle Re-Identification (Re-ID) aims to identify the same vehicle across different cameras, hence plays an important role in modern traffic management systems. The technical challenges require the algorithms must be robust in different views, resolution, occlusion and illumination conditions. In this paper, we first analyze the main factors hindering the Vehicle Re-ID performance. We then present our solutions, specifically targeting the dataset Track 2 of the 5th AI City Challenge, including (1) reducing the domain gap between real and synthetic data, (2) network modification by stacking multi heads with attention mechanism, (3) adaptive loss weight adjustment. Our method achieves 61.34% mAP on the private CityFlow testset without using external dataset or pseudo labeling, and outperforms all previous works at 87.1% mAP on the Veri benchmark. The code is available at https://github.com/cybercore-co-ltd/track2_aicity_2021., Comment: Accepted to CVPR Workshop 2021, 5th AI City Challenge

Optimal hydrophone performance for 1-3 piezoelectric composites is achieved from the design of material properties. The piezocomposite consists of piezoceramic rods immersed in a polymer matrix. We obtain the effective moduli of the piezocomposite by the differential effective medium theory and the results are explicitly dependent on volume fractions of the piezoceramic rods and elastic properties of the matrix. Topology optimization with level set method is used to optimize the elastic properties of the matrix phase. Numerical examples propose three-dimensional microstructures with negative Poisson’s ratio for the polymer matrix that can enhance hydrostatic charge coefficients of the piezocomposite.

In this work, we present an efficiently computational approach for designing material micro-structures by means of topology optimization. The central idea relies on using the isogeometric analysis integrated with the parameterized level set function for numerical homogenization, sensitivity calculation and optimization of the effective elastic properties. Design variables, which are level set values associated with control points, are updated from the optimizer and represent the geometry of the unit cell. We further improve the computational efficiency in each iteration by employing reduced order modeling when solving linear systems of the equilibrium equations. We construct a reduced basis by reusing computed solutions from previous optimization steps, and a much smaller linear system of equations is solved on the reduced basis. Two- and three-dimensional numerical results show the effectiveness of the topology optimization algorithm coupled with the reduced basis approach in designing metamaterials.

We present two new dead pixel correction algorithms for dual-band infrared imagery. Specifically, we address the problem of repairing unresponsive elements in the sensor array using signal processing techniques to overcome deficiencies in image quality that are present following the nonuniformity correction process. Traditionally, dead pixel correction has been performed almost exclusively using variations of the nearest neighbor technique, where the value of the dead pixel is estimated based on pixel values associated with the neighboring image structure. Our approach differs from existing techniques, for the first time we estimate the values of dead pixels using information from both thermal bands collaboratively. The proposed dual-band statistical lookup (DSL) and dual-band inpainting (DIP) algorithms use intensity and local gradient information to estimate the values of dead pixels based on the values of unaffected pixels in the supplementary infrared band. The DSL algorithm is a regression technique that uses the image intensities from the reference band to estimate the dead pixel values in the band undergoing correction. The DIP algorithm is an energy minimization technique that uses the local image gradient from the reference band and the boundary values from the affected band to estimate the dead pixel values. We evaluate the effectiveness of the proposed algorithms with 50 dual-band videos. Simulation results indicate that the proposed techniques achieve perceptually and quantitatively superior results compared to existing methods.

Summary Background Haemopoietic stem-cell transplantation (HSCT) eradicates host haemopoiesis before venous infusion of haemopoietic stem cells (HSCs). The pathway to cellular recovery has been difficult to study in human beings because of risks associated with interventions during aplasia. We investigated whether 18 F-fluorothymidine ( 18 F-FLT) imaging was safe during allogenic HSCT and allowed visualisation of early cellular proliferation and detection of patterns of cellular engraftment after HSCT. Methods Eligible patients were aged 18–55 years, had high-risk haematological malignancies. All patients underwent myeloablation followed by HSCT. The imaging primary endpoint was detection of early subclinical engraftment after HSCT with 18 F-FLT PET or CT. Imaging was done 1 day before and 5 or 9, and 28 days, and 1 year after HSCT. This study is registered with ClinicalTrials.gov, number NCT01338987. Findings Between April 1, 2014, and Dec 31, 2015, 23 patients were enrolled and assessable for toxic effects after completing accrual. 18 F-FLT was not associated with any adverse events or delayed engraftment. 18 F-FLT imaging objectively identified subclinical bone-marrow recovery within 5 days of HSC infusion, which was up to 20 days before engraftment became clinically evident. Quantitatively, 18 F-FLT intensity differed significantly between myeloablative infusion before HSCT and subclinical HSC recovery (p=0·00031). 18 F-FLT biodistribution over time revealed a previously unknown path of cellular recovery of haemopoiesis in vivo that mirrored fetal ontogeny. Interpretation 18 F-FLT allowed quantification and tracking of subclinical bone-marrow repopulation in human beings and revealed new insights into the biology of HSC recovery after HSCT. Funding National Institutes of Health, Ben's Run/Ben's Gift, Albert and Elizabeth Tucker Foundation, Mex Frates Leukemia Fund, Jones Family fund, and Oklahoma Center for Adult Stem Cell Research.

In this paper we present an implementation of the two-stage time-integration technique for solving semi-discrete systems of elastodynamics problems. This numerical approach effectively suppress the spurious high-frequency oscillations without affecting the accuracy of low modes, especially for long-term integration. At the stage of basic computations zero-dissipative Runge-Kutta methods are applied which yield no error accumulation due to numerical dissipation in the obtained solutions. Then for the suppression of spurious high-frequency oscillations, the filtering stage including pre- or/and post-processing are proceeded by using time integration methods with large numerical dissipation. Therein, the minimum necessary amount of numerical dissipation for filtering the spurious oscillations are determined by using a calibration procedure in which the amplitudes of spurious oscillations occurring the solution profile are minimized by varying stepsizes. The two-stage time integration technique is implemented...

Clinical assessment of bone marrow is limited by an inability to evaluate the marrow space comprehensively and dynamically and there is no current method for automatically assessing hematopoietic activity within the medullary space. Evaluating the hematopoietic space in its entirety could be applicable in blood disorders, malignancies, infections, and medication toxicity. In this paper, we introduce a CT/PET 3D automatic framework for measurement of the hematopoietic compartment proliferation within osseous sites. We first perform a full-body bone structure segmentation using 3D graph-cut on the CT volume. The vertebrae are segmented by detecting the discs between adjacent vertebrae. Finally, we register the bone marrow CT volume with its corresponding PET volume and capture the spinal bone marrow volume. The proposed framework was tested on 17 patients, achieving an average accuracy of 86.37% and a worst case accuracy of 82.3% in automatically extracting the aggregate volume of the spinal marrow cavities.

Current noninvasive evaluation of bone marrow proliferation in leukemia treatment is limited to manually examining marrow tissue in multiple regions of interest (ROIs). The statistics extracted from these ROIs often fail to provide an accurate global characterization of the patient's marrow. We propose an automatic framework for segmenting spinal marrow compartments to characterize the bone marrow from full-body joint PET/CT scans acquired subsequent to bone marrow transplantation. We first apply a graph-cut algorithm to the CT volume to obtain a 3D full-body bone map. We then isolate the spinal column in a single sagittal plane where connected components labeling and iterative thresholding are used to segment the vertebral bodies. This fully automated approach achieves an average accuracy of 91.7% and a worst case accuracy of 80.4% in testing on 51 scans of 17 patients. Finally, we outline a method for rejecting the cortical bone in transverse planes that can be combined with the sagittally segmented vertebral bodies to obtain a 3D map of the vertebral body medullary cavities for the entire spine.

We propose a new polynomial-time grayscale conversion algorithm applicable to general color images. The output grayscale image is modeled as a linear combination of three color channels where the mixing coefficients are computed by a constrained quadratic programming scheme using modulation domain features of the input color image. The optimization is formulated such that local color space distances between pixels in the input image are approximately preserved in the output grayscale image. Our experimental results demonstrate that the proposed method performs favorably compared to state-of-the-art conversion algorithms, often producing a grayscale image with better visual distinctions between patches that are close in the color space of the original image.

We introduce an SIR particle filter for tracking civilian targets including vehicles and pedestrians in dual-band midwave/longwave infrared imagery as well as a novel dualband track consistency check for triggering appearance model updates. Because of the paucity of available dual-band data, we constructed a custom sensor to acquire the test sequences. The proposed algorithm is robust against magnification changes, aspect changes, and clutter and successfully tracked all 17 cases tested, including two partial occlusions. Future work is needed to comprehensively evaluate performance of the algorithm against state-of-the-art video trackers, especially considering the relatively small number of previous dual-band tracking results that have appeared.

As the demand for wind energy increases, industry and policymakers have been pushing to place larger wind turbines in denser wind farms. Furthermore, there are higher expectations for reliability of turbines, which require a better understanding of the complex interaction between wind turbines and the fluid flow that drives them. As a test platform, we used the Whisper 500 residential scale wind turbine to support structural and atmospheric modeling efforts undertaken to improve understanding of these interactions. The wind turbine’s flexible components (blades, tower, etc.) were modeled using finite elements, and modal tests of these components were conducted to provide data for experimental validation of the computational models. Finally, experimental data were collected from the wind turbine under real-world operating conditions. The FAST (Fatigue, Aerodynamics, Structures, and Turbulence) software developed at the National Renewable Energy Laboratory was used to predict total system performance in terms of wind input to power output along with other experimentally measurable parameters such as blade tip and tower top accelerations. This paper summarizes the laboratory and field test experiments and concludes with a discussion of the models’ predictive capability. LA-UR-12-24832.

We proposed an automatic nonlinear texture-cartoon decomposition based on the frequency behavior of texture and cartoon across different scales. We measured the ratio of gradient magnitude across modulation domain components and used this ratio to classify the texture and cartoon pixels. The algorithm computed the modulation domain component where texture and cartoon are separated. Our simulation results showed that the proposed algorithm is able to extract meaningful texture and cartoon components from images efficiently.

We introduce a novel decomposition algorithm capable of extracting locally coherent and visually meaningful texture components from images. The algorithm estimates texture dominant orientation for each coherent component and iteratively extracts it from the image based on a new quantitative coherency measure formulated in the modulation domain. The original image is perfectly reconstructed from extracted components.

For the first time, we demonstrate modulation domain image filters that achieve perceptually motivated image processing goals by directly manipulating the FM functions in a multi-component AM-FM image model. The action of previous modulation domain filters has been limited to modification of the AM functions based on the values of the AM and FM functions. This is because reconstruction of the modified phase from the filtered frequency modulation vectors was an unsolved problem. Here, we present two new algorithms capable of reconstructing the phase from the processed frequencies, one based on a least squares solution of the discrete Poisson equation with Neumann boundary condition and one based on cubic tensor product spline integration. New modulation domain FM filters are designed to modify both the orientations and magnitudes of the visually important emergent image frequency vectors. In our most dramatic example, we demonstrate an FM filter that autonomously changes the stripes on the pants in the well known Barbara image from vertical to horizontal.

We combine an adaptation of the steerable image pyramid sub- band decomposition with a spline-based perfect reconstruction demodulation algorithm to obtain an invertible AM-FM image transform. For the first time, we achieve perceptually-based signal processing goals by applying filtering operations directly to the computed subband amplitude and frequency modulations. The results are dramatic and would be difficult or impossible to obtain by linear processing. In our most interesting example, a simple AM-FM filter succeeds in smoothly and naturally removing the bands from the hat in the well-known Lena image.

Challenging infrared data sequences such as the well-known AMCOM closure sequences are characterized by highly nonstationary evolutionary target and clutter signatures, poor target-to- clutter ratios, and complex kinematics arising from both the target motion and the motion of the sensor platform itself. In such cases, track consistency checks can provide a valuable means for detecting an imminent track loss. In this paper, we consider a simple target model with a correlation-based detection process and a straightforward SIR particle filter track processor. We show that the performance of the track processor can be dramatically improved by incorporating modulation domain consistency checks to identify failure in the correlation-based detection process. This strategy results in a robust dual-domain tracker that, despite the simplicity of its state model, delivers superior tracking performance against the very difficult AMCOM sequences.

For the first time, we perform normalized correlation template tracking in the modulation domain. For each frame of the video sequence, we compute a multi-component AM-FM image model that characterizes the local texture structure of objects and backgrounds. Tracking is carried out by formulating a modulation domain correlation function in the derived feature space. Using visible and longwave infrared sequences as illustrative examples, we study the performance of this new approach relative to two basic pixel domain correlation template trackers. We also present preliminary results from a new dual domain tracker that operates simultaneously in both the pixel and modulation domains.

The Electrophysics PV320 is a broadband thermal imaging system with several attractive features including low cost (about USD 25K including optics and software), small size, uncooled operation with a BST sensor array, spectral response from 0.6 to 14 μm, easily interchangeable warm optics, and on board USB 2.0 digital video output. In this paper we describe the technical challenges that were involved in integrating together two copies of the PV320L2Z camera variant to create an experimental dual-band IR data acquisition system for measuring targets, backgrounds, and clutter. The PV320 manufacturer-supplied software includes a user friendly, all-in-one application as well as software development kits providing camera control routines that are callable from C++, Visual Basic, and LabView. While this software works well for operating a single PV320 camera, it does not provide any direct support for simultaneously imaging with multiple cameras. The main technical issues are that the base software driver can connect to only one camera at a time and that multiple instances of the driver cannot be loaded simultaneously. Therefore, to achieve our goal of acquiring dual-band IR signatures, it was necessary to program a custom distributed algorithm capable of running two copies of the driver simultaneously on two separate computers with one PV320L2Z connected to each.

We compute joint AM-FM models that characterize infrared targets and backgrounds in the modulation domain. We consider spatially localized structures within an IR image as sums of nonstationary, quasi-sinusoidal functions admitting locally narrowband amplitude and frequency modulations. By quantitatively estimating the modulations that dominate the signal spectrum on a spatially local basis, we obtain a new modulation domain feature vector that can augment the more traditional pixel domain, Fourier spectrum, and multispectral color features that have been used in IR target detection and tracking systems for a long time. Our preliminary studies, based primarily on midwave and longwave missile approach sequences, suggest that IR targets and backgrounds do typically possess sufficient spatially local modulated structure (i.e., texture) for modulation domain techniques to be meaningfully applied. We also present qualitative results strongly indicating that the modulation domain feature vector is a powerful tool for discriminating infrared targets and backgrounds.

Nanoparticles (NPs) and surfactants can spontaneously concentrate at the interface between two immiscible liquids, such as oil and water. Systems of high oil–water interfacial area, such as emulsions, are the basis of many industries and consumer products. Although NPs and surfactants are currently incorporated into many of these applications, their mutual interfacial behavior is not completely understood. Here we present molecular dynamics simulations of NPs and non-ionic surfactant in the vicinity of an oil–water interface. It was found that in low concentration the surfactants and NPs show cooperative behavior in lowering the oil–water interfacial tension, while at higher surfactant concentration this synergy is attenuated. It was also found that binding of surfactants to the NP surface decreases the surfactant efficiency in lowering the interfacial tension, while concurrently creating a barrier to NP aggregation.