id
stringlengths 20
32
| text
stringlengths 100
121k
| image_path
stringlengths 15
235
|
|---|---|---|
arxivcap_2012.05361_2 | Title: Data-driven Competitive Algorithms for Online Knapsack and Set Cover | Caption: Threshold function $\Psi_\alpha$ in Eq. (<ref>) with different values of $\alpha$ for ; $\gamma = 20$. Degradation factor {$\left(\okpalg(\alpha)\right)$} for different value of $\gamma$ for . | 2012.05361_2.jpg |
arxivcap_2108.09597_6 | Title: Hierarchical Summarization for Longform Spoken Dialog | Caption: Alg 1. Semantic Segmentation Example of one speaker turn input into our coreference resolution algorithm. On the left, the coreference tags generated from the AllenNLP coreference resolution module are shown with six different references highlighted. Our algorithm groups sentences with references into semantic chunks with a minimum limit of references and words so that the semantic chunks are still meaningful. In Semantically Clustered Chunk #1 (blue), the first sentence includes three different references (1,2,3), of which two (2,3) are still used in the second sentence, hence why it is included. The singleton contains no references and is segmented out. In Semantically Clustered Chunk #2 (red), the first sentence contains two references (4,5) of which one (4) is in the second sentence. Although there is a new reference (6), the reference terminates within the second sentence so no further sentences are added. | 2108.09597_6.jpg |
arxivcap_1401.8253_5 | Title: Evaluating the Applicability of the Fokker-Planck Equation in Polymer Translocation: A Brownian Dynamics Study | Caption: Mean translocation time, $\langle\tau_1\rangle$, vs $N-N_{\rm p}$, for several different values of pore monomer friction coefficient, $\gamma_{0{\rm p}}$. The polymer starts midway in the pore, with $s_0=(N-1-n_{\rm p})/2$. In addition, $L=3$, $R=0.65$, and $f_{\rm d}=0$. The inset shows the translocation time scaling exponent, $\alpha$, vs $\gamma_{0{\rm p}}$. | 1401.8253_5.jpg |
arxivcap_2202.00274_2 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: Time evolution of the propagator in the HRW model ($p_0=p_{\max}=1$, $\sigma=1$, and $R(p)=|p|^2$). Symbols with lines are the numerical results of the HRW model by simulating trajectories of random walkers. The solid line represents a part of a steady-state solution, $\rho^*(p) \propto |p|^{-\alpha}$, for reference. The dashed lines represent plateaus around $|p|= 0$, which shift up with time $t$. Initial momentum is chosen uniformly on $[-1,1]$. | 2202.00274_2.jpg |
arxivcap_1702.06296_17 | Title: Gaia eclipsing binary and multiple systems. Supervised classification and self-organizing maps | Caption: Interpolated-scaled-shifted light curves from {CALEB}, ordered by their coefficients on LD1 (leftmost panel), the distribution of the system morphology classes by bins of LD1 scores (second panel from left), and the same for {Gaia}-sampled {Kepler} eclipsing binaries. In the panels showing the light curves, offsets were added for the sake of visibility. In the barplot panels, dark grey corresponds to detached systems, mid-grey, to semi-detached, near-contact and contact ones, and light grey, to overcontact ones. | 1702.06296_17.jpg |
arxivcap_2303.14737_6 | Title: Growing Convex Collision-Free Regions in Configuration Space using Nonlinear Programming | Caption: A comparison of the convex collision-free regions generated without (left) and with (right) an additional gripper orientation constraint designed to prevent a held mug from spilling its contents. The red pillars are task-space obstacles that both the robot and mug must not collide with. | 2303.14737_6.jpg |
arxivcap_1912.05088_2 | Title: An Introduction to Complex Systems Science and its Applications | Caption: Representative complexity profiles for random, coherent, and correlated systems (see <ref>). Any given system may have aspects of each at various scales. | 1912.05088_2.jpg |
arxivcap_1104.3616_4 | Title: Strategies used as spectroscopy of financial markets reveal new stylized facts | Caption: Average returns $R$ versus average holding time for individuals in A-share (a), institutions in A-share (b), individuals in B-share (c) and institutions in B-share (d), respectively. The symbols present the average values over all investors ($\circ$), as well as investors who earn positive return ({{$\vartriangle$}}) and negative return ({{$\triangledown$}}). The dashed line delineates the benchmark in absolute terms of zero return. The solid lines with error bars are the average simulation results of random trading. The insets in panels (c) and (d) are magnifications of the curves for small values of $\Delta t$. | 1104.3616_4.jpg |
arxivcap_1906.05317_4 | Title: COMET: Commonsense Transformers for Automatic Knowledge Graph Construction | Caption: Model diagram. (a) In the multi-headed attention module, the key, value, and query all pass through a head-specific projection before a scaled dot-product attention is computed between them. The outputs of the heads are concatenated and projected. (b) Inside the transformer block, the outputs of all the previous layer blocks from earlier time steps are input to the multi-headed attention with the preceding block for the current time step as the query. (c) Each token is an input to a first-layer block along with all preceding tokens. Dotted lines indicate outputs to all future blocks in the next layer and inputs from all preceding blocks in the previous layer. | 1906.05317_4.jpg |
arxivcap_2108.09671_2 | Title: Pi-NAS: Improving Neural Architecture Search by Reducing Supernet Training Consistency Shift | Caption: Ranking correlations on 792 architectures on NAS-Bench-201 <cit.> on CIFAR-10 without skip connection and zero operations compared to SPOS <cit.>, arch2vec <cit.> and ProxylessNAS <cit.>. | 2108.09671_2.jpg |
arxivcap_2303.14737_7 | Title: Growing Convex Collision-Free Regions in Configuration Space using Nonlinear Programming | Caption: A bimanual environment consisting of two KUKA LBR iiwa and a shelving unit. IRIS-NP can scale well to this 14 dimensional environment, constructing some regions in as little as 1 minute. | 2303.14737_7.jpg |
arxivcap_1401.8253_6 | Title: Evaluating the Applicability of the Fokker-Planck Equation in Polymer Translocation: A Brownian Dynamics Study | Caption: Translocation time distributions for unforced translocation where the polymer starts midway in the nanopore, with $s_0=(N-1-n_{\rm p})/2$. In addition, $L=3$, $R=0.65$, and $f_{\rm d}=0$. The open circles denote BD simulation data, and the solid curves show FP theoretical predictions. The inset shows $N_{\rm eff}$ vs $\gamma_{0{\rm p}}$, for $N_{\rm eff}$ values that provide the best match between the FP predictions and BD data. The solid line is a guide for the eye. | 1401.8253_6.jpg |
arxivcap_2202.00274_3 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: Time evolution of the propagator, i.e. data from Fig. <ref>, multiplied by $t^{1-\gamma}$ in the HRW model for different times ($\alpha =2$, $c=1$, $p_0=p_{\max}=1$, and $\sigma^2 =1$). Symbols with lines are the results of numerical simulations for the HRW model. The dashed lines represent the infinite density, i.e., Eq. (<ref>). The solid lines represent rescaled scaling functions, $t g_{\rm exp} (t^\gamma p)$. The dotted lines represent $t g_{\rm exp} (0)$ for different values of $t$. The initial momentum is chosen uniformly on $[-1,1]$. | 2202.00274_3.jpg |
arxivcap_1912.05088_3 | Title: An Introduction to Complex Systems Science and its Applications | Caption: The complexity profile of a factory that can produce a large number of copies of a few types of goods, and the complexity profile of a factory that can produce many types of goods but not in large numbers. The number of copies of a good produced is a proxy for scale since, given a fixed technology, mass production requires larger-scale coordinated action in the factory (e.g. an assembly line), and the number of different types of goods that can be produced at a given scale is a proxy for the number of different possible behaviors of the factory - and thus its complexity - at that scale. | 1912.05088_3.jpg |
arxivcap_2303.14655_1 | Title: GOAL: A Challenging Knowledge-grounded Video Captioning Benchmark for Real-time Soccer Commentary Generation | Caption: Distributions of the knowledge triples (left) and word number (right) over the video clip length. | 2303.14655_1.jpg |
arxivcap_2012.05361_3 | Title: Data-driven Competitive Algorithms for Online Knapsack and Set Cover | Caption: Degradation factor {$\left(\oscalg(\theta)\right)$} with different values of $m$ for Normalized cost of $\oscalg(\theta)$ for two different problem scenarios | 2012.05361_3.jpg |
arxivcap_2212.13254_1 | Title: Modeling high-entropy transition-metal alloys with alchemical compression | Caption: Different interpretations of the alchemical compression scheme. (a) In a conventional density-correlation ML scheme, each type of atoms is associated with a separate density. (b) The entries in the alchemical compression matrix $\Ualchemical$ can be interpreted as describing the "character" of each physical element in terms of $\Nalchemical$ pseudoelements - a concept that is not dissimilar from the notion of "classical elements". (c) The structure can be also seen as described in terms of a density of pseudo-elements, for which each site contains a contribuition from each of the compressed channels. | 2212.13254_1.jpg |
arxivcap_2108.09671_3 | Title: Pi-NAS: Improving Neural Architecture Search by Reducing Supernet Training Consistency Shift | Caption: Ranking correlations of diverse NAS methods. Values within the parentheses are their corresponding Kendall's Tau. Our $\Pi$-NAS gains the best correlation, indicating that $\Pi$-NAS effectively reduces the training consistency shift.(CP: cross-path learning; MT: mean teacher; DS: downsample-sharing; DA: learning different augmented views of the same image) | 2108.09671_3.jpg |
arxivcap_2202.13828_6 | Title: Detecting the power spectrum turnover with HI intensity mapping | Caption: Constraints on $H_0$ from a model-independent logarithmic polynomial fit to turnover scale of the power spectrum. By obtaining a dilation parameter, $\alpha_0$, (as discussed in {sec:StandardRulersIM}) an inference on $H_0$ can be obtained from the physics of matter-radiation equality. Black vertical dotted line marks the fiducial $H_0$ quantity and shaded regions indicate the $1\sigma$ confidence regions for both IM surveys tested. We have made the simplification that all other parameters are fixed, thus this represents an early demonstration rather than a robust forecast. | 2202.13828_6.jpg |
arxivcap_1906.05317_6 | Title: COMET: Commonsense Transformers for Automatic Knowledge Graph Construction | Caption: The percentage of novel ConceptNet development set tuples per minimum edit distance from training tuples. In green: classifier-scored accuracy of each subset. | 1906.05317_6.jpg |
arxivcap_2202.01518_1 | Title: Ammonia mapping observations toward the Galactic massive star-forming region Sh 2-255 and Sh 2-257 | Caption: (a) Three color composite image of Sh 2-255 and Sh 2-257. Blue, green, and red show DSS2 blue, (DSS2 blue + DSS2 red)/2, and DSS2 red, respectively. (b) $^{12}$CO $J=$2-1 integrated intensity map obtained by <cit.>. Purple and yellow cross indicates the position of the embedded cluster S255 IR and S255 N, respectively. The yellow boxes show the mapping area of our observation. {Blue dotted circles enclose the H {II} regions. The size of H {II} regions are obtained from <cit.> and <cit.>.} | 2202.01518_1.jpg |
arxivcap_2303.14655_3 | Title: GOAL: A Challenging Knowledge-grounded Video Captioning Benchmark for Real-time Soccer Commentary Generation | Caption: The distribution of sentence-level knowledge density in terms of relation types and knowledge triples. | 2303.14655_3.jpg |
arxivcap_2108.09671_4 | Title: Pi-NAS: Improving Neural Architecture Search by Reducing Supernet Training Consistency Shift | Caption: Ranking correlations on all 15625 architectures on NAS-Bench-201 <cit.> on CIFAR-10 compared to SPOS <cit.>, arch2vec <cit.> and ProxylessNAS <cit.>. | 2108.09671_4.jpg |
arxivcap_1401.8253_7 | Title: Evaluating the Applicability of the Fokker-Planck Equation in Polymer Translocation: A Brownian Dynamics Study | Caption: Distribution of the pore emptying time $\tau-\tau_1$ for unforced translocation. In this stage, the polymer starts at $s=0$ at time $\tau_1$ and proceeds until the time $\tau$, at which the nanopore is fully empty. The monomer pore friction is $\gamma_{0\rm p}=50$. In addition, $N=21$, $L=3$ and $R=0.65$. The open circles denote BD simulation data, and the solid curve shows the FP theoretical prediction. The inset shows the free energy curve used in the calculation. The blue circles indicates the two possible starting points of the pore emptying stage. | 1401.8253_7.jpg |
arxivcap_2202.13828_7 | Title: Detecting the power spectrum turnover with HI intensity mapping | Caption: Constraints on the turnover scale for a single-realisation of simulated SKA-MID data with foreground contamination, cleaned using PCA. The blue-circle data points show how we have successfully corrected for signal loss with a transfer function, which even assumed an incorrect cosmology. Orange-square data points show the effects from signal loss where no transfer function is used. The $1\sigma$ confidence regions on the estimated turnover scale are displayed by the vertical shaded colour regions, which show how using the transfer function allows an unbiased recovery of the fiducial turnover scale (black-dotted vertical line.) | 2202.13828_7.jpg |
arxivcap_1912.05088_4 | Title: An Introduction to Complex Systems Science and its Applications | Caption: Schematic complexity profiles of militaries in conflict. Top: If two armies are operating with the same number of possible behaviors but at different scales, the larger-scale one is favored. Middle: If two armies are operating at the same scale but with different numbers of possible behaviors, the higher-complexity one is favored. Bottom: If two armies are operating at different scales and with different numbers of possible behaviors, which one is favored depends on the terrain (see text). Note that these profiles are simplified to highlight the key concepts; actual militaries operate at multiple scales. More generally, the top and middle graphs depict conflicts in which one army has at least as much complexity as the other at every scale. | 1912.05088_4.jpg |
arxivcap_2202.00274_4 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: Rescaled propagator of the HRW model for different values of $\sigma^2$ ($\alpha =2$, $c=1$, $p_0=p_{\max}=1$, and $t=10^4$). Symbols with lines are the results of numerical simulations for the HRW model. The dashed solid line represents the scaling function, i.e., Eq. (<ref>). The initial position is chosen uniformly on $[-1,1]$. Note that the results for different $\sigma^2$ are indistinguishable. | 2202.00274_4.jpg |
arxivcap_1906.05428_3 | Title: Simulating quantum thermodynamics of a finite system and bath with variable temperature | Caption: Temperatures $T_\mathcal{E}$ converge to the Einstein solid temperature relation as the number of bath oscillators $\eta \to \infty$. Deviations outside this limit are due to the finite size of the bath. | 1906.05428_3.jpg |
arxivcap_2303.14655_5 | Title: GOAL: A Challenging Knowledge-grounded Video Captioning Benchmark for Real-time Soccer Commentary Generation | Caption: Comparison of conventional video captioning with its knowledge-grounded setting. A soccer commentary describe the fine-grained entities ({red}), their relevant structural ({blue}), and semi-/unstructured knowledge ({green}). | 2303.14655_5.jpg |
arxivcap_2212.13254_2 | Title: Modeling high-entropy transition-metal alloys with alchemical compression | Caption: Learning curves for different models. Full lines correspond to models built using only $V^{(\text{aeb})}$ and $V^{(\text{3B})}$, with $\Nalchemical$ pseudo-elements (all optimized iteratively). The dotted green curve that is obtained with a $\Ualchemical$ filled with uniform random numbers. The dashed green line corresponds to a model that includes $V^{(\text{aeb})}$ and $V^{(\text{3B})}$, as well as the full set of pair potentials and a non-linear term built on top of the contracted power spectrum features $V^{(\text{NN})}$. | 2212.13254_2.jpg |
arxivcap_2202.13828_8 | Title: Detecting the power spectrum turnover with HI intensity mapping | Caption: Constraints on the turnover scale for a single-realisation of simulated SKA-MID data including a frequency-dependent cosine beam with side-lobes. The blue-circle data points show how we have successfully corrected for the beam when using the correct model. Orange-square data points show the effects from where a very incorrect beam model is used which assumed a Gaussian frequency-independent beam without side-lobes. The $1\sigma$ confidence regions on the estimated turnover scale are displayed by the vertical shaded colour regions, which show how an incorrect beam model can cause a biased recovery of the fiducial turnover scale (black-dotted vertical line.) | 2202.13828_8.jpg |
arxivcap_2108.09671_6 | Title: Pi-NAS: Improving Neural Architecture Search by Reducing Supernet Training Consistency Shift | Caption: Ranking correlations on 4096 architectures on NAS-Bench-201 <cit.> on CIFAR-10 without skip connection operation compared to SPOS <cit.>, arch2vec <cit.> and ProxylessNAS <cit.>. | 2108.09671_6.jpg |
arxivcap_2202.01518_2 | Title: Ammonia mapping observations toward the Galactic massive star-forming region Sh 2-255 and Sh 2-257 | Caption: NH$_3$ spectra at A $(l,b)=(\timeform{192.613D}, \timeform{-0.045D})$, B $(l,b)=(\timeform{192.572D}, \timeform{-0.045D})$, and C $(l,b)=(\timeform{192.634D}, \timeform{-0.159D})$. These positions are also presented in Figure <ref>(a) | 2202.01518_2.jpg |
arxivcap_2012.05325_3 | Title: Convolutional Neural Networks for Multispectral Image Cloud Masking | Caption: Example showing the RGB false color composite and the cloud mask obtained with the $33\times33$ {patch to pixel} CNN. | 2012.05325_3.jpg |
arxivcap_2202.00274_5 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: EB parameter as a function of $\gamma$ ($=1/\alpha$) for the kinetic energy, i.e., ${\mathcal O}(p)=p^2$. Symbols are the results of numerical simulations for the HRW, deterministic, and exponential models. The solid line represents ${\rm A}(\gamma)$ and ${\rm ML}(\gamma)$ for $\gamma<1/3$ and $\gamma>1/3$, respectively. The dashed line represents Eq. (<ref>). The solid line represents Eq. (<ref>) and (<ref>). | 2202.00274_5.jpg |
arxivcap_2303.14655_6 | Title: GOAL: A Challenging Knowledge-grounded Video Captioning Benchmark for Real-time Soccer Commentary Generation | Caption: The distribution of the top-$50$ relations in our dataset. | 2303.14655_6.jpg |
arxivcap_2202.13828_9 | Title: Detecting the power spectrum turnover with HI intensity mapping | Caption: Robustness of the turnover scale $k_0$ to intensity mapping observational effects. We show a model power spectrum decomposed into 2-dimensional modes perpendicular and parallel to the line-of-sight ($k_\perp,k_\parallel$). The highest intensity region is where the power spectrum peaks and is thus the turnover region. We mark this exact position with the turnover scale $k_0$ shown by the red dashed contour. In each panel we show the impact to the power spectrum from different observational effects including; signal loss due to foreground cleaning, a smoothing to the field caused by the telescope beam, and the presence of scale-independent thermal noise. The final panel shows the combination from all observational effects. | 2202.13828_9.jpg |
arxivcap_1401.8253_8 | Title: Evaluating the Applicability of the Fokker-Planck Equation in Polymer Translocation: A Brownian Dynamics Study | Caption: Translocation time distributions for driven translocation for several values of $\gamma_{0{\rm p}}$. The symbols denote BD simulation data and the solid curves show theoretical predictions using the FP formalism. The data are for a polymer of length $N=41$ starting at $s_0=2.04$ and subject to a driving force of $f_{\rm d}=1$. In addition, the pore dimensions are $L=3$ and $R=0.65$. The inset shows $N_{\rm eff}$ vs $\gamma_{0{\rm p}}$. These values of $N_{\rm eff}$ provided the best match between the FP and BD distributions. | 1401.8253_8.jpg |
arxivcap_2108.09656_5 | Title: ExamGAN and Twin-ExamGAN for Exam Script Generation | Caption: Training exam script vs. generated exam scripts in terms of the four properties (M1-Difficulty, M2-Validity, M3-Rationality, M4-Distinguishability) | 2108.09656_5.jpg |
arxivcap_2202.01518_3 | Title: Ammonia mapping observations toward the Galactic massive star-forming region Sh 2-255 and Sh 2-257 | Caption: Integrated intensity map of (a) NH$_3$ (1,1), (b) NH$_3$ (2,2), and (c) NH$_3$ (3,3). The lowest contour level and intervals are 0.09 K km s$^{-1}$ ($\sim 3\sigma$) and 0.12 K km s$^{-1}$($\sim 4\sigma$), respectively. A, B, and C represent the positions of the spectra in Figure <ref>. The purple, yellow crosses, and blue dotted circles are also the same as Figure <ref>(b). | 2202.01518_3.jpg |
arxivcap_2305.05384_4 | Title: Generation of imprinted strain gradients for spintronics | Caption: (a) - (b) effective strain ($\epsilon_{xx}-\epsilon_{yy}$) and and strain gradient ($\frac{d}{dx}\left( \epsilon_{xx}-\epsilon_{yy} \right)$) for, respectively, a square and a triangular pair of opening obtained from FEM simulations. The $x$ axis refers to the red dashed line. (c) - (d) surface strain $\epsilon_{xx}-\epsilon_{yy}$ obtained with FEM simulations. (e) - (f) in-plane hysteresis loops obtained with Kerr microscopy on a full film of Co$_{70}$Fe$_{30}$ (30 nm) for, respectively, a square and a triangular pair of opening. The contrast was measured before the opening (diamond) and between them (star) according to the marker position. The magnetic field was applied along the direction $\Phi=0^\circ$. | 2305.05384_4.jpg |
arxivcap_1912.05088_5 | Title: An Introduction to Complex Systems Science and its Applications | Caption: Complexity profiles of two hierarchies, each with the same number of people. Here, the scale is the number of coordinated man-hours. In one hierarchy, all decisions, regardless of the scale, are made by a single person, while in the other, different decisions are made at various levels of the hierarchy. | 1912.05088_5.jpg |
arxivcap_2012.05458_4 | Title: Beyond Class-Conditional Assumption: A Primary Attempt to Combat Instance-Dependent Label Noise | Caption: Training/testing accuracy and critical sample ratio throughout training on IDN and CCN with varying noise fractions. | 2012.05458_4.jpg |
arxivcap_1906.05428_5 | Title: Simulating quantum thermodynamics of a finite system and bath with variable temperature | Caption: Histogram of total quantum state probabilities per unit energy for an initial Gaussian state (a) and corresponding time-evolved equilibrium state (b) with a random matrix coupling with $k=0.0027$. The total probability per unit energy does not converge to zero at high energy for the equilibrium state, indicating a problem with the coupling. | 1906.05428_5.jpg |
arxivcap_2303.14750_1 | Title: An Approach for Generating Families of Energetically Optimal Gaits from Passive Dynamic Walking Gaits | Caption: A demonstration of our approach on (a) a two-link biped robot with an actuator at the hip using (b) the passive (i.e., unactuated) dynamic walking motion of the biped model to compute continuous sets of energetically optimal actuated gaits, including (c) gaits that walk on level ground. Using a single passive gait as a seed value from a family of unactuated gaits (green curve), we are able to generate a curve of gaits with the same walking speed across a range of slopes (black curve in gray plane) and then switch to generating a curve of gaits that walk on level-ground across a range of walking speeds (black curve in yellow plane). The model is scaled with time measured in units of $t_0$ and speed in units of $v_0$ (see Table <ref>). | 2303.14750_1.jpg |
arxivcap_2212.13254_4 | Title: Modeling high-entropy transition-metal alloys with alchemical compression | Caption: Parity plot between reference energy and forces and the values computed with the model, for a hold-out set of 500 structures, randomly selected from the training set. Energy error: 10 meV/atom mean absolute errror (MAE), 14 meV/atom root mean square error (RMSE), Force error: 190 meV/Å MAE, 280 meV/Å RMSE. | 2212.13254_4.jpg |
arxivcap_2202.13828_10 | Title: Detecting the power spectrum turnover with HI intensity mapping | Caption: Sky area required to achieve detection ($\alpha{>}0$) of the turnover feature in the power spectrum for a MeerKAT survey in L-band and UHF-band. Shaded areas represent the $1\sigma$ and $2\sigma$ confidence regions on $\alpha$ measurements. Thin black vertical line indicates the sky area currently proposed for a MeerKLASS survey <cit.>. Far-left panel assumes a constant time per pointing ($t_\text{p}$) for each area, hence why the total observation time scales proportionally with sky area (shown by top $x$-axis). This keeps noise at a constant level as area changes. Other panels show different cases of fixed total observation time (displayed in panel titles), all lower than that currently proposed for MeerKLASS (4,000$\,$hrs). In these panels, noise increases with increasing area, which is why constraints at large areas begin to worsen. | 2202.13828_10.jpg |
arxivcap_2108.09656_13 | Title: ExamGAN and Twin-ExamGAN for Exam Script Generation | Caption: Questions in training exam scripts vs. questions in generated exam scripts. | 2108.09656_13.jpg |
arxivcap_2202.01586_4 | Title: Backscatter-Aided NOMA V2X Communication under Channel Estimation Errors | Caption: The effect of channel estmation error on the total achievable energy efficiency of V2XneT by varying $P_{max}$. | 2202.01586_4.jpg |
arxivcap_2303.14753_1 | Title: Does "Deep Learning on a Data Diet" reproduce? Overall yes, but GraNd at Initialization does not | Caption: Correlation between GraNd at Initialization and Input Norm for CIFAR-10's training set. ({fig:grand_init_vs_input_norm:original}, {fig:grand_init_vs_input_norm:hlb}, {fig:grand_init_vs_input_norm:joost}): We sort the samples by their average normalized score (i.e., the score minus its minimum divided by its range), plot the scores and compute Spearman's rank correlation on CIFAR-10's training data. The original repository and the `minimal' implementation have very high rank correlation - `hlb' has a lower but still strong rank correlation. ({fig:grand_init_div_input_norm:joost}): Ratio between input norm and gradient norm. In the `minimal' implementation, the ratio between input norm and gradient norm is roughly log-normal distributed. | 2303.14753_1.jpg |
arxivcap_2202.00274_6 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: Time evolution of the propagator multiplied by $t^{\gamma-1}$ in the deterministic model for different times ($\alpha=\gamma^{-1} =2, c=1$, and $p_{\rm trap}=1$). Symbols with lines represent the results of numerical simulations of the deterministic model. The dashed lines represent the infinite invariant density $I_\infty(p)$ given by Eq. (<ref>). The solid lines represent rescaled scaling functions, $t g_{\rm det} (t^\gamma p)$. The dotted lines represent $t g_{\rm det} (0)$ for different values of $t$. The initial position is chosen uniformly on $[-1,1]$. | 2202.00274_6.jpg |
arxivcap_2012.05302_2 | Title: Generative Adversarial Networks for Annotated Data Augmentation in Data Sparse NLU | Caption: This schematic represents our data and experimental process. For our experiments we use annotated data collected by <cit.>. Train and validation sets were combined together in a single train data set to extract pseudo Golden utterances. Pseudo Goldens were further used to train and evaluate the GAN model. The test set was untouched and was used to evaluate model performance in full scale. To simulate the low resource feature task, we used the full train data set for the two robust domains, and the pseudo Goldens to represent the third low-resource domain. | 2012.05302_2.jpg |
arxivcap_1906.05428_6 | Title: Simulating quantum thermodynamics of a finite system and bath with variable temperature | Caption: Time evolved state with the "tamed" coupling Eq. <ref> has probabilities that converge to zero at high energy. The initial state was the same as panel (a) of Fig. <ref>. | 1906.05428_6.jpg |
arxivcap_2108.09663_2 | Title: SIDE: Center-based Stereo 3D Detector with Structure-aware Instance Depth Estimation | Caption: Network architecture of the proposed . It takes only the stereo RGB images as inputs. After the backbone network, the network outputs 7 branches which predict center depth, 2D center heatmap, stereo 2D boxes, object's kepoints, dimension and steering angle of 3D box based on the idea of anchor-free. Center depth branch is different from others, in which we perform instance depth estimation by selecting 2D RoI of objects. In order to get accurate depth, we introduce match-reweight and structure-aware attention in the depth estimation to make depth feature more focused. Finally, detection results are further refined through the post-processing. | 2108.09663_2.jpg |
arxivcap_1912.05088_6 | Title: An Introduction to Complex Systems Science and its Applications | Caption: A complexity profile of a system with a separation of scales. A separation of scales implies that the behaviors occurring below a certain scale ($s_0$ in the above figure) are at larger scales mostly independent from one another and that therefore, at these larger scales, only the average effects of the small-scale behaviors are relevant. | 1912.05088_6.jpg |
arxivcap_2202.13736_1 | Title: On the Robustness of CountSketch to Adaptive Inputs | Caption: Left: Bias-to-noise ratio for number of rounds, average of 10 simulations with different initializations (shaded region between the minimum and maximum). Right: Attack rounds versus sketch size $\ell=d/b$ to obtain bias-to-noise ratio $\in \{1,4\}$, averaged over $20$ and $5$ simulations respectively (shaded region between minimum and maximum). | 2202.13736_1.jpg |
arxivcap_1401.8253_10 | Title: Evaluating the Applicability of the Fokker-Planck Equation in Polymer Translocation: A Brownian Dynamics Study | Caption: Translocation time distributions for driven translocation for several values of the driving force, $f_{\rm d}$. The symbols denote BD simulation data and the solid curves show theoretical predictions. The data are for a polymer of length $N=41$ starting at $s_0=2.04$ and a pore with a friction of $\gamma_{0{\rm p}}=70$ and dimensions of $L=3$ and $R=0.65$. The FP curves were calculated using $N_{\rm eff}=3.35$. The inset shows $\langle\tau_1\rangle$ and $\langle\tau\rangle$ vs $f_{\rm d}^{-1}$. The solid lines are fits to the data using a power law functional form. The fits yielded a scaling of $f_{\rm d}^{-0.984\pm 0.006}$ for both $\langle\tau_1\rangle$ and $\langle\tau\rangle$. | 1401.8253_10.jpg |
arxivcap_2212.13254_5 | Title: Modeling high-entropy transition-metal alloys with alchemical compression | Caption: MAE for the formation energy of binary compounds from the Materials Project database. The inset shows a representative hull plot for the Ti-Pt system, highlighting the hulls obtained from the single-point DFT calculations and the ML predictions. The dashed line identifies the structures that are stable based on the energies available in the Materials Project database. | 2212.13254_5.jpg |
arxivcap_2305.05384_5 | Title: Generation of imprinted strain gradients for spintronics | Caption: Details of the FEM simulations performed to obtain the stress/strain distribution. Stress tensor on a Si/SiOx wafer simulated using COMSOL Multiphysics after deposition of a tensile SiN over-layer. (a) stress tensor on the surface of the wafer which is uniform. (b) surface displacement of a 5" wafer. (c) discretization and (d) x component of the strain tensor simulated in the proximity of a removed compressive SiN layer. | 2305.05384_5.jpg |
arxivcap_2202.01487_1 | Title: A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes | Caption: PSDS_1 results for systems selected for scenario 1, evaluated with reference evaluation set and sets with varying TNTSNR. | 2202.01487_1.jpg |
arxivcap_1802.01192_3 | Title: MOCCA-SURVEY Database I: Assessing GW kick retention fractions for BH-BH mergers in globular clusters | Caption: Example of how each recoil velocity component depends on $q$. For $v_m$ mass ratio is the only variable, so we have a deterministic formula. Other components and the overall kick velocity depend also on spin magnitudes and orientations, in this case we plot the mean value. As we can see, the major component is almost always $v_\parallel$, others only play a role for low $q$. This plot was done for MS1 mass - spin relationship (e. g. Section <ref>) | 1802.01192_3.jpg |
arxivcap_2012.05302_3 | Title: Generative Adversarial Networks for Annotated Data Augmentation in Data Sparse NLU | Caption: Data quality evaluation for synthetic data generated in different GAN models: number of unique utterances generated normalized by the number of Goldens used as an input for GAN model (left), and 4-gram BLEU score (right). | 2012.05302_3.jpg |
arxivcap_1906.05428_7 | Title: Simulating quantum thermodynamics of a finite system and bath with variable temperature | Caption: Time-dependent temperatures $T_{\mathcal{SE}}(t)$ (Eq. <ref>) for a series of calculations with approximately the same energy $E \approx 6.14$ but different starting levels $n_0$. Each temperature evolves to approximately the same final temperature $T_{\mathcal{SE}} \approx 1.92$ from Eq. <ref>. | 1906.05428_7.jpg |
arxivcap_2108.09663_4 | Title: SIDE: Center-based Stereo 3D Detector with Structure-aware Instance Depth Estimation | Caption: The red dot represents the left boundary of the visible range, and the green dot represents the right boundary of the visible range. It can be seen from the figure that the prediction of the visible range is inaccurate for cars with severe occlusion. | 2108.09663_4.jpg |
arxivcap_1912.05088_7 | Title: An Introduction to Complex Systems Science and its Applications | Caption: A normal distribution (thin-tailed) and a distribution with a power-law decay (fat-tailed). The fat-tailed distribution may appear more stable, due to the lower probability of small-scale fluctuations and the fact that samples from the distribution may not contain any extreme events. However, sooner or later, a fat-tailed distribution will produce an extreme event, while one could wait thousands of lifetimes of the universe before a normal distribution produces a similarly extreme event. Note that the axes of this graph are truncated; the illustrated fat-tailed distribution can, with small but non-negligible probability (0.04%), produce events with a scale of one million or more. | 1912.05088_7.jpg |
arxivcap_1702.06296_20 | Title: Gaia eclipsing binary and multiple systems. Supervised classification and self-organizing maps | Caption: Relative populations of the system morphology classes of {CALEB} in the 2-dimensional reduced space of the SOM. Red: overcontact, violet: semidetached, near-contact and contact, green: detached binaries. The area of the plotted piechart is proportional to the fraction of objects belonging to the cell in the full population. | 1702.06296_20.jpg |
arxivcap_2202.00274_7 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: Rescaled propagators for different distributions $\chi(p)$ ($\alpha=\gamma^{-1} =2$, $c=1$, and $p_{\rm trap}=1$), where we consider the uniform distribution $\chi(p)=1/2$ on $p\in [-1,1]$ and the Gaussian distribution $\chi (p) = \exp(-p^2/2)/\sqrt{2\pi}$. Symbols with lines are the results of the numerical simulations of the deterministic model with $t=10^4$. The solid line represents the scaling function given by Eq. (<ref>). The initial position is chosen uniformly on $[-1,1]$. Note that the results for different $\chi(p)$ are indistinguishable. | 2202.00274_7.jpg |
arxivcap_2202.01487_2 | Title: A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes | Caption: PSDS_2 results for systems selected for scenario 1, evaluated with reference evaluation set and sets with varying TNTSNR. | 2202.01487_2.jpg |
arxivcap_2108.09663_5 | Title: SIDE: Center-based Stereo 3D Detector with Structure-aware Instance Depth Estimation | Caption: Depth estimation module builds a 4D cost volume and performs 3DCNN to estimate the depth of a 3D bounding box center. We use two modules Match Reweight and Structure-Aware Attention to make information more focused and thus beneficial for estimating the depth. | 2108.09663_5.jpg |
arxivcap_2012.05302_4 | Title: Generative Adversarial Networks for Annotated Data Augmentation in Data Sparse NLU | Caption: Schematic representation of the reward function produced by discriminator model for each of the $n+1$ input tokens and N Monte-Carlo rollouts {(avg = average)}. Tokens fixed at each iteration are represented by solid squares and tokens produced by the rollout policy are represented by blank squares. Sentence-level reward produces feedback on a complete sentence only, whereas token-level reward gives feedback on each token conditioned on previous set of tokens. | 2012.05302_4.jpg |
arxivcap_2202.01487_3 | Title: A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes | Caption: PSDS_2 results for systems selected for scenario 2, evaluated with reference evaluation set and sets with varying TNTSNR. | 2202.01487_3.jpg |
arxivcap_1401.8253_11 | Title: Evaluating the Applicability of the Fokker-Planck Equation in Polymer Translocation: A Brownian Dynamics Study | Caption: Translocation time distributions $g(\tau_1)$ for driven translocation for polymers of several lengths. In each case translocation starts at $s_0=3.44$. In addition, $\gamma_{0\rm p}=70$ and $f_{\rm d}=1$. The symbols denote BD simulation data and the solid curves show theoretical predictions. The FP curves were calculated using $N_{\rm eff}=3.35$. The inset shows $\langle\tau_1\rangle$ vs $\Delta s$, where $\Delta s=N-1-n_{\rm p}-s_0$. The solid line is a fit to the data using a power law. The fit yielded a scaling of $\langle\tau_1\rangle \propto \Delta s^{1.025\pm 0.004}$. | 1401.8253_11.jpg |
arxivcap_2305.05384_8 | Title: Generation of imprinted strain gradients for spintronics | Caption: (a) surface strain ($\epsilon_{xx}-\epsilon_{yy}$) for a single square-shaped opening obtained from FEM simulations. (b) - (d) angular plot of the normalized remanent magnetization $M_r/M_s$ as a function of the in-plane angle $\Phi$ of the magnetic field measured with Kerr microscopy on a full film of Co$_{40}$Fe$_{40}$B$_{40}$ (30 nm) where a single opening in the SiN is present. The measurements were taken in different locations around the etched area in the passivation layer as indicated by the markers. Namely, an area far from the opening (b), on the left side where $\epsilon_{xx}-\epsilon_{yy}>0$ (c) and on the lower side where $\epsilon_{xx}-\epsilon_{yy}<0$ (d) are considered. | 2305.05384_8.jpg |
arxivcap_2212.13254_6 | Title: Modeling high-entropy transition-metal alloys with alchemical compression | Caption: Equation of state for the random relaxed (RR) and fully relaxed (FR) structures (see text for the full definition), computed with the potential and with the reference DFT. Birch-Murnaghan parameters for cohesive energy ($E_0$), equilibrium volume ($V_0$), bulk modulus ($B_0$), bulk modulus derivative ($B'_0$) are given in the table. | 2212.13254_6.jpg |
arxivcap_1906.05428_8 | Title: Simulating quantum thermodynamics of a finite system and bath with variable temperature | Caption: System level probabilities evolve in time to the Boltzmann distribution at temperature $T_{\mathcal{SE}}(E=\langle \hat H \rangle)$ from Eq. <ref>. {The decay of the initial state $n_0=0$ is described by Eq. <ref> with $\tau = 1.02 \pm 0.03$ and $\delta = 2.38 \pm 0.06$.} | 1906.05428_8.jpg |
arxivcap_2202.13855_2 | Title: Large-Scale 3D Semantic Reconstruction for Automated Driving Vehicles with Adaptive Truncated Signed Distance Function | Caption: Our large-scale 3D reconstruction, texturing and semantic mapping pipeline for automated driving vehicles using LiDAR and camera sensors. An accurate 3D geometric environment model is obtained from accumulated LiDAR point clouds using Adaptive TSDF. The reconstructed 3D model is accurately textured from a series of registered camera images with an optimal image patch selection strategy and color optimization techniques. With a MRF-based data fusion approach, the 3D geometric model is then semantic mapped using multiple semantic segmented camera images. | 2202.13855_2.jpg |
arxivcap_1912.05088_8 | Title: An Introduction to Complex Systems Science and its Applications | Caption: A representative complexity profile of a complex system. Understanding all the details (i.e. all of the small-scale behaviors) is impossible and unnecessary; the most important information is contained in the large-scale behaviors. However, for systems for which mean-field theory does not apply, characterizing these behaviors will involve more than a simple average. | 1912.05088_8.jpg |
arxivcap_1702.06296_21 | Title: Gaia eclipsing binary and multiple systems. Supervised classification and self-organizing maps | Caption: Relative populations of the system morphology classes of {Kepler} systems (bottom) in the 2-dimensional reduced space of the SOM. Colour code is the same as for Figure <ref>. The area of the plotted piechart is proportional to the fraction of objects belonging to the cell in the full population, except for $n<3$ in the bottom panel, which were inflated to the size of $n=3$ for visibility. | 1702.06296_21.jpg |
arxivcap_2202.00274_8 | Title: Infinite ergodic theory for three heterogeneous stochastic models with application to subrecoil laser cooling | Caption: EB parameter as a function of $\gamma$ for two observables ${\mathcal O}(p)=p^2$ and ${\mathcal O}(p)=I(|p|>0.5)$, where ${\mathcal O}(p)=I(|p|>0.5)=1$ if $|p|>0.5$ and zero otherwise. The solid line represents ${\rm A}(\gamma)$ and ${\rm ML}(\gamma)$ for $\gamma<1/3$ and $\gamma>1/3$, respectively. The dashed line represents ${\rm ML}(\gamma)$ for $\gamma<1/3$. Note that $I(|p|>0.5)$ is integrable with respect to $I_\infty (p)$ for all $\gamma$. | 2202.00274_8.jpg |
arxivcap_1802.01192_4 | Title: MOCCA-SURVEY Database I: Assessing GW kick retention fractions for BH-BH mergers in globular clusters | Caption: GW kick retention fraction as a function of $\sigma$ (deviation of the Gaussian for drawing $\theta$ in primordial case). Top: MS0, middle: MS1, bottom: MS2. Regardless of the model for reasonable values of $\sigma$ ($10^{\circ}-20^{\circ}$) the plot saturates around the value of $0.3$. Therefore the exact value of $\sigma$ assumed doesn't have the crucial effect on the GW kick retention fractions (although it may influence other results). At low $\sigma$ fraction dependence on $\sigma$ is bigger, because $\sigma\sim1^\circ$ impies almost perfect alignment and therefore low kicks ($v_{\parallel}\approx0$, and $v_{\parallel}$ is usually the biggest component, e. g. Fig. <ref>). We also plot GW kick retention fractions for primordial only and dynamical only. The latter displays a small dependence on $\sigma$ even though $\sigma$ is only important for $\theta$ choice in primordial binaries. This is due to the fact that initial spin magnitudes for dynamical mergers in some cases depend on previous primordial mergers, which in turn depend on $\sigma$. However this is a secondary effect with no serious scientific implications. | 1802.01192_4.jpg |
arxivcap_2202.01487_4 | Title: A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes | Caption: PSDS_1 results for systems selected for scenario 1, evaluated with synthetic set with varying onset time. | 2202.01487_4.jpg |
arxivcap_2108.09603_2 | Title: Tensor Pooling Driven Instance Segmentation Framework for Baggage Threat Recognition | Caption: Difference between conventional structure tensors (used in <cit.>), and proposed multi-scale tensor approach. First row shows the original scans from OPIXray <cit.>, GDXray <cit.>, and SIXray <cit.> dataset. The second row shows the output for the conventional structure tensors <cit.>. The third row shows the output for the proposed tensor pooling module. | 2108.09603_2.jpg |
arxivcap_2012.05377_4 | Title: Pump efficacy in a fluid-structure interaction model of a chain of contracting lymphangions | Caption: Velocity field surrounding the central valve (valve 3) at select times during cycle 10 in a 4lymphangion, fastorthograde pump with an AAPD of 0.051 cmH$_2$O. Every other velocity vector is included in the plot, the vectors are scaled by a factor of 0.125, and the (unscaled) velocity magnitude is plotted with color bars measured in mm/s | 2012.05377_4.jpg |
arxivcap_2202.01472_2 | Title: Lattice QCD calculation of the two-photon exchange contribution to the muonic-hydrogen Lamb shift | Caption: Theoretical results for the TPE contribution. From top to bottom, the results are refered to Refs. <cit.>,<cit.>,<cit.>,<cit.>,<cit.> and <cit.>, respectively. | 2202.01472_2.jpg |
arxivcap_1806.08737_5 | Title: Constraining the multiplicity statistics of the coolest brown dwarfs:
binary fraction continues to decrease with spectral type | Caption: Same as Figure <ref> for the extended sample combining our observed survey with the $\geqslant$T8 brown dwarfs selected from <cit.> and <cit.>. Contours denote the 0%, 50% and 100% completeness regions. The red star shows the position of the T8.5$+$T9 binary discovered in <cit.>, W0458$+$6434AB, using the component masses calculated in this work. The yellow star corresponds to the known W0146$+$4234 binary, unresolved in our observations. | 1806.08737_5.jpg |
arxivcap_2212.13254_7 | Title: Modeling high-entropy transition-metal alloys with alchemical compression | Caption: Comparison between the potential energy evaluated along two 10ps MD/MC trajectories, and that recomputed by DFT for 100 snapshots. The inset shows the parity plot for the force components computed for those structures. Energies have a MAE of 14 (48) meV/atom and forces a component MAE of 0.23 (0.29) eV/Å for the 300 (5000) K trajectory. | 2212.13254_7.jpg |
arxivcap_1912.05088_9 | Title: An Introduction to Complex Systems Science and its Applications | Caption: An illustration from ref. <cit.> showing the interplay between cooperation and competition in the context of sports teams and leagues. | 1912.05088_9.jpg |
arxivcap_1906.05428_9 | Title: Simulating quantum thermodynamics of a finite system and bath with variable temperature | Caption: Time-averaged equilibrium system probabilities for three initial states (a), (b), and (c) with the energies and temperatures in Table <ref>. The Boltzmann distributions $\rho_\mathcal{S}^{n,n} \sim \exp({-E_n/T_\mathcal{SE}})$ at the analytical temperatures $T_\mathcal{SE}$ give very good descriptions of the system level probabilities $\rho_\mathcal{S}^{n,n}$, while the Boltzmann distributions at the infinite bath $T$ do not. | 1906.05428_9.jpg |
arxivcap_1401.8030_1 | Title: Transit Fare Arbitrage: Case Study of San Francisco Bay Area Rapid Transit (BART) System | Caption: Fare as a function of distance traveled. Center grey line is the price proportional to distance. The blue line above the grey line penalizes short distances while the red line below subsidizes them. Fig <ref> attempts to explain these curves using population density. | 1401.8030_1.jpg |
arxivcap_2305.05391_2 | Title: Privacy-preserving Adversarial Facial Features | Caption: Reconstructed images from adversarial features with different noise bounds on datasets LFW, CFP-FP, and AgeDB-30. | 2305.05391_2.jpg |
arxivcap_2202.13855_12 | Title: Large-Scale 3D Semantic Reconstruction for Automated Driving Vehicles with Adaptive Truncated Signed Distance Function | Caption: MRF in comparison to naive approach. | 2202.13855_12.jpg |
arxivcap_1702.06296_22 | Title: Gaia eclipsing binary and multiple systems. Supervised classification and self-organizing maps | Caption: Confusion matrices obtained by imposing the best-fit boundaries in the SOM space of CALEB to separate system morphology classes. Left panel: confusion matrix of {CALEB} objects; middle panel: confusion matrix of estimated classes by imposing the same boundaries on the KIC objects in the SOM space; right panel: average confusion matrix obtained by applying the 500 {CALEB}-trained Random Forest models to {Kepler} objects. The numbers in the white boxes are absolute numbers, those in the orange boxes are fractions (bottom row: contamination rate in the estimated class, rightmost column: error rate). | 1702.06296_22.jpg |
arxivcap_1806.08737_6 | Title: Constraining the multiplicity statistics of the coolest brown dwarfs:
binary fraction continues to decrease with spectral type | Caption: Same as Figure <ref> for the sample of T5$-$T7.5 sample selected from <cit.> and <cit.>. Contours denote the 0%, 50% and 100% completeness regions. The red star shows the position of T5$+$T5 binary discovered in <cit.>, W1841$+$7000AB. | 1806.08737_6.jpg |
arxivcap_2202.00198_2 | Title: Recognition-Aware Learned Image Compression | Caption: Sample images from BPG and our model, trained here with $\lambda=0.9, \beta=1.0$, initial learning rate of 5e-5, and dropout of 0.2. (a) has BPP=0.132, PSNR=21.55; (b) has BPP=0.119, PSNR=17.62; (c) has BPP=0.117, PSNR=26.35; (d) has BPP=0.119, PSNR=18.36. | 2202.00198_2.jpg |
arxivcap_1401.8053_1 | Title: Hallucinating optimal high-dimensional subspaces | Caption: The projection matrix $\mathbf{P} \in \mathbb{R}^{25 \times 100}$ modelling the process of downsampling a $10 \times 10$ pixel image to $5 \times 5$ pixels, using (a) bilinear and (b) bicubic projection models, shown as an image. For the interpretation of image intensities see the associated grey level scales on the right. | 1401.8053_1.jpg |
arxivcap_2108.09603_3 | Title: Tensor Pooling Driven Instance Segmentation Framework for Baggage Threat Recognition | Caption: Contour instance segmentation from multi-scale tensors. The first column shows the original scans, the second column shows the multi-scale tensor representations, the third column shows the ground truths, and the fourth column shows the extracted contours of the contraband items. | 2108.09603_3.jpg |
arxivcap_2012.05377_11 | Title: Pump efficacy in a fluid-structure interaction model of a chain of contracting lymphangions | Caption: Pump-function plots for chains of one to four lymphangions with different contraction styles. Plots are grouped into panels (a) through (e) according to contraction style. The data-point labelling and linear regression are as in Fig. <ref>. The legend applies to panels (a)-(e), and LA in the legend denotes lymphangion | 2012.05377_11.jpg |
arxivcap_2202.01472_5 | Title: Lattice QCD calculation of the two-photon exchange contribution to the muonic-hydrogen Lamb shift | Caption: Results of $\Delta E_{\lat}$ as a function of the integral range $R$ at $\{\Delta t_i,\Delta t_f,t_s\}=\{2a,2a,4a\}$. The upper and lower panels show the results for the connected and disconnected contribution, respectively. Results from different terms have been slightly shifted for clarity. | 2202.01472_5.jpg |
arxivcap_1912.05016_1 | Title: Robust Feature-Based Point Registration Using Directional Mixture Model | Caption: The process of preparing surface normals from pointcloud data. 1) pointcloud 2) estimating surface normals 3) outlier removal 4) spherical clustering. | 1912.05016_1.jpg |
arxivcap_1906.05384_2 | Title: Loop Programming Practices that Simplify Quicksort Implementations | Caption: Pictorial example of how loop rotation coalesces duplicate code. Legend: blue are loop statements, green are non-loop statements. The statements "{BO}", "{PR}", "{UI}", and "{EX}" are duplicated between pre-loop and loop. They may be coalesced by a loop rotation. {fig:ferris-wheel-dup-node}: Original program; {fig:ferris-wheel-rotated}: Rotated program. | 1906.05384_2.jpg |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.