id
stringlengths 20
32
| text
stringlengths 100
121k
| image_path
stringlengths 15
235
|
|---|---|---|
arxivcap_2202.01537_3 | Title: Bending Graphs: Hierarchical Shape Matching using Gated Optimal Transport | Caption: The result of dense shape matching on MPI-FAUST<cit.> trained on a few samples of SURREAL<cit.>. (a) we have a target shape to which we are finding correspondences. Coarse matches (b) show robust matched patches to the target shape. In (c) we visualize the final confidence values predicted. The maps suggest more confidence in less deformed regions. In (d) we can see our fine and dense correspondences by passing our coarse matches to the FM algorithm in form of corresponding delta functions. Finally, we have the Ground correspondence map in (e) and error map in (f) which highlights our good performance for inter-subject cases without the need for refinement. | 2202.01537_3.jpg |
arxivcap_2305.05360_1 | Title: Anisotropic hybridization probed by polarization dependent x-ray absorption spectroscopy in VI3 van der Waals Mott ferromagnet | Caption: VI$_3$ crystal structure and orbital order. Side (a) and top (b) view of the hard spheres representation of a single layer VI$_3$. (c) Perspective views of the VI$_6$ cluster unit and the relative $3d$ energy levels splitting due to a trigonal Jahn-Teller deformation (compression and elongation occur along the trigonal, i.e. 3-fold, axis). (d) Perspective view of the VI$_3$ low temperature monoclinic unit cell; the $c$ axis is along the [0001] direction perpendicular to the vdW planes. In all the illustrations, V and I atoms are depicted by purple and yellow spheres respectively. Principles of XNLD and XMCD techniques. (e) In XNLD, XAS spectra are acquired with horizontal (blue) and vertical (orange) linearly polarized x-rays and in absence of any magnetic field; in our experiment the x-ray beam impings on the sample at 20$^{\circ}$ grazing incidence. (f) In a XMCD experiment, left- (blue) and right- (orange) circularly polarized x-rays while the magnetic field $B$ is applied parallel to the x-ray beam; in our experiment the x-ray beam is normal to the sample. | 2305.05360_1.jpg |
arxivcap_1109.6664_1 | Title: Conditional pair distributions in many-body systems: Exact results for Poisson ensembles | Caption: (Color online) {Left:} Example of a point pattern with a characterization of the evaluation of the pair distribution functions $g(\omega)$, $g_v(\omega)$ and $g_{vv'}(\omega)$. The point $P'$ lies in the surface of the spherical volume $\omega$ centered at the reference point $P$. Nearest neighbors of $P$ and $P'$ lie in the surfaces of spherical volumes $v$ and $v'$ centered in $P$ and $P'$, respectively. {Right:} Representation of the four classes of ($v,v'$) pairs (see text). Here, $v_+\equiv \max(v,v')$ and $v_-\equiv \min(v,v')$. | 1109.6664_1.jpg |
arxivcap_1401.8257_6 | Title: Online Clustering of Bandits | Caption: Plots on the Yahoo datasets reporting Clickthrough Rate (CTR) over time, i.e., the fraction of times the algorithm gets payoff one out of the number of retained records so far. | 1401.8257_6.jpg |
arxivcap_1111.0261_9 | Title: Extended Lagrangian free energy molecular dynamics | Caption: (Color online) The fluctuations in the total free energy, $E^{\rm tot}_F = E_{\rm K} + U - T_e{\cal S}$, for a water molecule, H$_2$O (PBE0/3-21G), using either the exact Fermi-Dirac distribution or the recursive Fermi operator expansion, Algorithm <ref>, with 5 ($m=5$) or 6 ($m=6$) recursion steps. The simulation is based on a hybrid density functional (PBE0 <cit.>) for $T_e = 10,000$ K and $T_{\rm ion} \approx 300$ K using 2 SCF cycles per time step, $\delta t = 0.5$ fs, within the density matrix extension. | 1111.0261_9.jpg |
arxivcap_2212.13138_12 | Title: Large Language Models Encode Clinical Knowledge | Caption: Clinician evaluation of possible bias in answers Flan-PaLM answers were found to contain biased information in 7.9% of the cases. However, this number reduced to 0.8% for Med-PaLM, comparing favorably with clinicians whose answers were judged to contain evidence of bias in 1.4% of the cases. | 2212.13138_12.jpg |
arxivcap_2206.01897_2 | Title: Modeling of Textures to Predict Immune Cell Status and Survival of Brain Tumour Patients | Caption: The results of AUC-ROC and confusion matrix (left: Macrophage M1, middle: Neutrophils, right: T Cells Follicular Helper) obtained from RF models use the deep radiomic features to predict the immune cell status of 151 brain tumors patients (LGG+GBM). | 2206.01897_2.jpg |
arxivcap_1802.01329_1 | Title: Reshaping of a Janus ring | Caption: The dependence of quantitative characteristics of non-planar configurations with different looping numbers $n$ on the parameter $q$. (a): The energy decrements relative to the planar loop. (b): The mean square deviation from the original plane. | 1802.01329_1.jpg |
arxivcap_2303.14662_6 | Title: OTAvatar: One-shot Talking Face Avatar with Controllable Tri-plane Rendering | Caption: Qualitative result for multi-view reenactment. Examples are sampled from the multi-view dataset <cit.>. All methods use the first frame of the frontal-view portrait to extract the identity feature, and take the expressions of sequential frames and poses of different camera views to generate the talking face. Note that this subject is not included in the training set of any methods. | 2303.14662_6.jpg |
arxivcap_2212.13102_2 | Title: Piezoelectric effect and diode effect in anapole and monopole superconductors | Caption: The DOS in the $A_g + i B_{3u}$ anapole state with the following parameters: (a) $\alpha = +0.4$, $\beta = 1$, (b) $\alpha = -0.4$, $\beta = 1$, (c) $\alpha = +0.4$, $\beta = 0.7$, and (d) $\alpha = -0.4$, $\beta = 0.7$. We set $q_0 =0$ for simplicity. | 2212.13102_2.jpg |
arxivcap_1906.05253_3 | Title: Search on the Replay Buffer: Bridging Planning and Reinforcement Learning | Caption: Visual Navigation: We compare our method (SoRB) to prior work on the visual navigation environment (Fig. <ref>), using RGB images and depth images . We find that only our method succeeds in reaching distant goals. Baselines: SPTM <cit.>, C51 <cit.>, VIN <cit.>, HER <cit.>. | 1906.05253_3.jpg |
arxivcap_1912.05084_18 | Title: Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology | Caption: =10pt Results for a data set simulated according to the process detailed in Section <ref> in the Supplementary Material with sample size $n = 1000$, $q=2$ episodic components and $p=1$ regular components, each subject having $m_{i}=3$ replicates. From top to bottom, the left panels show the estimated densities $f_{X,\ell}(X_{\ell})$ of the two episodic components and the one regular component, respectively, obtained by our method (in blue) and the method of <cit.> (in red). The right panels show the estimated distributions of the scaled errors $f_{\epsilon,q+\ell}(\epsilon_{q+\ell})$ and the estimated variance functions $v_{\ell}(\wt{X}_{\ell}) = s_{\ell}^{2}(\wt{X}_{\ell})$, estimated by our method. The black lines represent the truths (the right panels) or their importance sampling based approximations (the left panels). | 1912.05084_18.jpg |
arxivcap_1104.3722_7 | Title: Investigating the Distribution of Password Choices | Caption: Entropy values for Uniform model, Real data and Zipf model. | 1104.3722_7.jpg |
arxivcap_1709.03144_29 | Title: The 12C(a,g)16O reaction and its implications for stellar helium burning | Caption: (Color online) Level diagram of the $^{16}$O compound nucleus. Levels that are irrelevant to the analysis are omitted. For example, unnatural parity states below the proton separation energy. Several reactions that populate the CN are shown where their energy axis has been converted to excitation energy. At low energy only $^{12}$C+$\alpha_0$, $^{16}$O+$\gamma$ and $^{16}$N($\beta\alpha$) partitions are considered. At higher energies, the $^{15}$N+$p$ and $^{12}$C+$\alpha_1$ partitions are also included. Representative experimental cross section measurements that populate the CN are shown on the right. The total cross section data for the $^{12}$C$(\alpha,\gamma)^{16}$O reaction are those of {schurmann2005}, the $^{15}$N$(p,\gamma_0)^{16}$O those of {PhysRevC.82.055804}, the $^{16}$N$(\beta\alpha)^{12}$C spectrum is from {PhysRevLett.70.726}, and the $^{12}$C$(\alpha,\alpha_0)^{12}$C data are from {PhysRevLett.88.072501}. The solid red curve represents the phenomenological $R$-matrix fit described in this work. | 1709.03144_29.jpg |
arxivcap_2108.09628_4 | Title: DisenKGAT: Knowledge Graph Embedding with Disentangled Graph Attention Network | Caption: An architecture overview of our DisenKGAT. The whole model contains three key modules: (1) relation-aware aggregation, (2) independence constraint, and (3) adaptive scoring. Each channel (color) represents a specific component and different components keep individuals with each other by the correlation restriction. The final prediction result depends on all components of the entities and is adaptive to the given scenario (relation). Best viewed in color. | 2108.09628_4.jpg |
arxivcap_2202.00257_1 | Title: Position-Dependent Snap Feedforward: A Gaussian Process Framework | Caption: Benchmark system for position-dependent snap feedforward, representing an H-type gantry commonly used in e.g. semiconductor back-end machines or large-format printing systems seen in the top right. The first mode shape of the flexible beam is displayed, with two actuators $a_1$ and $a_2$ and the output $y$. The performance location can vary, indicated by the scheduling variable $\rho$. | 2202.00257_1.jpg |
arxivcap_2005.06188_5 | Title: Star clusters near and far; tracing star formation across cosmic time | Caption: The distribution of star clusters onto the Galactic XY-plane. Blue dots represent open clusters and associations while red triangles mark GCs. The border of the Galactic disk is represented by the dotted circle (diameter 20 kpc). The local spiral arms are marked as thick solid cyan and magenta sections and are formally formally extended to the edge of the disk with thin solid curves. The light (yellow) thick circle around the Sun with radius of 1.8 kpc marks the completeness limit of the survey. The cross at (X, Y) = (8.5, 0) kpc indicates the position of the Galactic center. This figure was published as Fig.2 in <cit.>. | 2005.06188_5.jpg |
arxivcap_2206.11995_2 | Title: Efficient and Accurate Top-$K$ Recovery from Choice Data | Caption: F1 dataset ($n=22$): Average training time (seconds) against sample size for $m=2,4,8$. Choice-based Borda count ({ blue}) is several orders of magnitude faster than its competitors. | 2206.11995_2.jpg |
arxivcap_2012.05350_4 | Title: Automatic Diagnosis of Malaria from Thin Blood Smear Images using Deep Convolutional Neural Network with Multi-Resolution Feature Fusion | Caption: Training/Validation accuracy of DilationNet-C over epochs for resolution $(128\times128)$: The convergence of the DilationNet architecture is very fast. Validation accuracy achieves considerable value in a few numbers of training epochs. | 2012.05350_4.jpg |
arxivcap_1806.08917_11 | Title: Halide-Perovskite Resonant Nanophotonics | Caption: Integration of halide perovskites with non-perovskite nanostructures. (a) An embedded dielectric/metal heterostructure for plasmonic output of dielectric laser. Left: schematic illustration. Central: {electron} image in false colors. Right: optical images in reflection and luminescent modes <cit.>. (b) Left: {electron} image of the MAPbBr$_3$ microwire on silicon grating, where the scale bar is 10 $\mu$m. Right: fluorescent microscope image <cit.>. (c) Schematic showing the time-delayed pump excitations at 800 and 400 nm, which photogenerate free carriers in MAPbI$_3$ and MAPbBr$_3$, respectively <cit.>. | 1806.08917_11.jpg |
arxivcap_1902.04067_11 | Title: Multi-tier Caching Analysis in CDN-based Over-the-top Video Streaming Systems | Caption: A flowchart illustrates the online updates for an edge-cache when a file $i$ is requested at time $t_i$. Here, $t_{j,\ell}$ represents the time of the last request of file $i$ from edge-cache $j$, and $\mathcal{H}$ is the index set of all video files in the edge-cache. | 1902.04067_11.jpg |
arxivcap_2202.01537_4 | Title: Bending Graphs: Hierarchical Shape Matching using Gated Optimal Transport | Caption: We process deformable shape pairs by two levels of hierarchy, local graphs and shape graphs. Such hierarchical structure offers a flexible and holistic shape representation that enables correspondence matching and provides rich features for an optimal transport matching stage. | 2202.01537_4.jpg |
arxivcap_2205.02281_6 | Title: Three-Body Problem in Modified Dynamics | Caption: Upper panels: The time evolution ($300~\mathcal{T}$) of a binary system, each with mass $0.1~\mathcal{M}$, and a third object with mass $0.01~\mathcal{M}$. The initial distance of the third particle is $4.5~\mathcal{D}$ on the $y$ axis. The Middle panels: The initial distance of the third particle is decreased to $1.5~\mathcal{D}$ on the $y$ axis. Lower panels: The initial distance of the third particle is reduced to $1.0~\mathcal{D}$. The irregularity of the orbits increases from top to bottom; however, the lower MOD panel is still bounded while Newtonian evolution shows an ejection. The coordinates $x$ and $y$ are in units of $\mathcal{D}$. | 2205.02281_6.jpg |
arxivcap_2004.03425_10 | Title: Elemental abundances in M31: [Fe/H] and [{\alpha}/Fe] in M31 Dwarf Galaxies Using Coadded Spectra | Caption: [Fe/H] vs [$\alpha$/Fe] for our ten M31 dSphs, ordered from most massive to least massive. Small grey circles indicate abundance measurements of individual stars. Square symbols show validation coadds, and diamond symbols indicate science coadds. Fitted slopes and their corresponding $1\sigma$ uncertainties measured from individual measurements are shown in grey, and the slopes from all coadds (validation and science combined) for a given dSph are shown in magenta. | 2004.03425_10.jpg |
arxivcap_2206.11995_3 | Title: Efficient and Accurate Top-$K$ Recovery from Choice Data | Caption: Irish-Meath dataset ($n=14$): Exact top-$K$ accuracy against sample size. choice-based Borda count ({blue}) is competitive with baseline algorithms. Using larger menu sizes generally improves the performance of the algorithms. | 2206.11995_3.jpg |
arxivcap_2303.06773_6 | Title: Protecting Classical-Quantum Signals in Free Space Optical Channels | Caption: Fidelities obtained for a loss channel for the different non-Gaussian states. In each case, the squeezing of the TMSV state before the non-Gaussian operation is fixed at a value $r=4.7$ dB. | 2303.06773_6.jpg |
arxivcap_1802.01329_3 | Title: Reshaping of a Janus ring | Caption: Dependences of the energy gain comparative to initial state (a) and the mean square deviation from the original plane (b) on the parameter $q$ for different twist numbers $m$. | 1802.01329_3.jpg |
arxivcap_1906.05253_4 | Title: Search on the Replay Buffer: Bridging Planning and Reinforcement Learning | Caption: SoRB vs SPTM: Our method and Semi-Parametric Topological Memory <cit.> differ in the policy used and how distances are estimated. We find that both methods learn comparable policies, but our method learns more accurate distances. See text for details. | 1906.05253_4.jpg |
arxivcap_1211.6964_3 | Title: On the Bauschinger effect in supercoooled melts under shear: results from mode coupling theory and molecular dynamics simulations | Caption: Linear plot of the stress-strain relation $|\sigma(|\gamma-\gamma_{0}|)|$ at fixed temperature and shear rate for various flow histories: starting from equilibrium (EQ, red line), after flow reversal in the steady state (S, purple), from the elastic regime (el, blue), and from the point of the stress overshoot (max, green). Reversal at $\gamma_w^\text{s}=0.4$ is in the steady state as seen from comparing with reversal at $\gamma_w^\text{s}=75$ (black data). Smooth lines result from the schematic model. | 1211.6964_3.jpg |
arxivcap_2305.05360_2 | Title: Anisotropic hybridization probed by polarization dependent x-ray absorption spectroscopy in VI3 van der Waals Mott ferromagnet | Caption: XNLD on VI$_3$. Experimental XAS and corresponding XNLD spectra of VI$_3$ measured across the V $L_{2,3}$ edges in the absence of any magnetic field and in grazing incidence (GI, 20$^\circ$) for the FM phase at 20 K (a) and the PM phase at 100 K, i.e. above the structural transition (b). (c) Calculated XAS and XNLD spectra at 20 K. Experimental and theoretical data are reported with dotted and solid lines respectively. Horizontal- and vertical-linear polarization XAS and XNLD signal are shown respectively with blue, orange and green colors. Experimental XNLD signals are multiplied by a scale factor 3 for visualization. | 2305.05360_2.jpg |
arxivcap_1111.0261_10 | Title: Extended Lagrangian free energy molecular dynamics | Caption: (Color online) The relative error in the entropy as a function of the number of expansion steps $m$ in the recursive Fermi operator expansion, Algorithm <ref>, for three different electronic temperatures $T_e$. | 1111.0261_10.jpg |
arxivcap_1503.06984_3 | Title: Stability of discrete-time switching systems with constrained switching sequences | Caption: Using the M-Path-Dependent lifts, we produce better estimates. The estimates produced by the T-Product lifts need not decrease monotonically as $T$ increase. Lower bounds computed from the relative accuracies of the CJSR estimates. | 1503.06984_3.jpg |
arxivcap_2206.11995_4 | Title: Efficient and Accurate Top-$K$ Recovery from Choice Data | Caption: Irish West dataset ($n=9$): choice-based Borda count ({blue}) performs very similarly to MNL-MLE ({gray}) while the performance of ASR ({orange}) may diverge from MNL-MLE/choice-based Borda count. | 2206.11995_4.jpg |
arxivcap_2212.13102_4 | Title: Piezoelectric effect and diode effect in anapole and monopole superconductors | Caption: The spectrum of Bogoliubov quasiparticles in the group velocity model for the $A_g + i B_{3u}$ anapole state. We show the energy dispersion along the $k_x$ axis by setting $k_z = 0$ and (a),(c) $k_y = 0$ or (b),(d) $k_y = \pi/4$. The spectrum with $k_y = 0$ is gapless. (a),(b) The solid lines represent the spectrum calculated with $\alpha_x = 0.4$, while (c),(d) the dashed lines are obtained with $\alpha_x = -0.4$. We assume $\alpha_y = 0$, $\beta = 1$, $T=0.01$ and set $q_0 = 0$. | 2212.13102_4.jpg |
arxivcap_2202.00194_2 | Title: The binding of atomic hydrogen on graphene from density functional theory and diffusion Monte Carlo calculations | Caption: Visualization of the difference of PBE and DMC densities sliced along the 110 lattice plane of the unit cell for the graphene sheet, $\Delta\rho_{gr}$, (top) and H adsorbed onto graphene, $\Delta\rho_{dgr+H}$, (bottom). The abscissa represents traversing the 110 plane in fractional coordinates, while the ordinate represents traversing the $c$ axis in fractional coordinates. Blue regions represent places where the PBE density is larger, while the gold color represents regions where the DMC density is larger. | 2202.00194_2.jpg |
arxivcap_1109.6664_2 | Title: Conditional pair distributions in many-body systems: Exact results for Poisson ensembles | Caption: (Color online) Contact values $g(a^+)$ of the PDF for attractive hard spheres in the TSP formalism based on Eq. (<ref>) (solid lines), are compared with exact values in the limit of zero density (triangles), Monte Carlo simulations <cit.> (circles), results of a perturbative theory <cit.> (dashed lines), and results derived from Eq. (<ref>) (dotted lines). Here $T^*=kT/\epsilon$ is the reduced temperature, $\rho=n d^3$ the reduced number density, and $\lambda$ the potential range in units of the particle diameter $d$ [i.e., $\xi=a(\lambda^3-1)$ and $a=4\pi d^3/3$ in Eq. (<ref>)]. | 1109.6664_2.jpg |
arxivcap_2005.06188_6 | Title: Star clusters near and far; tracing star formation across cosmic time | Caption: Cumulative GMC mass spectra normalised by the observed area for different galaxies (left; from ) and for different regions of the spiral galaxy M51 (right; from ). The differences in slopes and maximum masses in different galaxies and for different kinematic environments of a given galaxy suggest an environmental dependence of the GMC mass distribution. | 2005.06188_6.jpg |
arxivcap_2303.14662_7 | Title: OTAvatar: One-shot Talking Face Avatar with Controllable Tri-plane Rendering | Caption: The interpolation results of identity code. Our model can generate smooth-varying identity attribute beyond motion control. | 2303.14662_7.jpg |
arxivcap_2102.12804_6 | Title: Compact Groups of Galaxies in Sloan Digital Sky Survey and LAMOST Spectral Survey. II. Dynamical Properties of Isolated and Embedded Groups | Caption: The distributions of the isolated (red hatched) CGs, embedded (blue filled) CGs, and control sample of noncompact groups (grey open). Upper left: redshift. Upper right: richness. Lower left: group luminosity, $L_{\text{19.5}}$. Lower right: LOS velocity dispersion, $\sigma_{\text{LOS}}$. | 2102.12804_6.jpg |
arxivcap_2206.11995_16 | Title: Efficient and Accurate Top-$K$ Recovery from Choice Data | Caption: F1 dataset: Exact top $K$ accuracy against sample size. | 2206.11995_16.jpg |
arxivcap_2210.09459_4 | Title: Extensible Proxy for Efficient NAS | Caption: The design of Eproxy and the searchable components. Dotted line: The configurable components for Discrete Proxy Search. Green block: Trainable components. The configurability of Eproxy can be further utilized by DPS to target search spaces/downstream tasks. | 2210.09459_4.jpg |
arxivcap_1605.05294_7 | Title: Migration and Growth of Protoplanetary Embryos III: Mass and Metallicity Dependence for FGKM main-sequence stars | Caption: {Left}: The $\eta_{\dot M}$-${\dot M}_g$ correlation for three different stellar masses (yellow, purple, and green lines for $0.5$, $1$, and $2 M_\odot$, respectively). The black dashed line represents $\eta_{\dot M}=0.5$, and three colored dots represent the average value of the observed accretion rate for different stellar masses ($M_{\ast} = 0.5, 1$ and $2 \ M_{\odot}$) in <cit.>'s work. {Right}: $\eta_{\dot M}-M_\ast$ correlation for different $\eta_c$ (solid and dashed lines for $\eta_c=0$ and $\eta_c=1 $ respectively). The red and blue colors correspond to ${\dot M}_{cr} = 1 \times 10^{-7} M_\odot$ yr$^{-1}$ and $5 \times 10^{-8} M_\odot$ yr$^{-1}$, respectively. For solar-type stars, $\eta_{\dot M} \sim 30\%-50\%$ as an upper limit for $\eta_J$. | 1605.05294_7.jpg |
arxivcap_2206.01897_3 | Title: Modeling of Textures to Predict Immune Cell Status and Survival of Brain Tumour Patients | Caption: Log-rank test and KM estimator use deep radiomic features (R, $\mathbf{y}=$126 features), clinical variables (C: age and gender), immune cell markers (I: Macrophage M1, Neutrophils and T Cells Follicular Helper) and their combination (R+C+I) to compare survival between two predicted survival groups based on RF models. | 2206.01897_3.jpg |
arxivcap_1806.08917_12 | Title: Halide-Perovskite Resonant Nanophotonics | Caption: Integration of halide perovskites with resonant nanostructures. (a) Gold plasmonic nanoparticles placed into m-TiO$_2$ layer of PSC. (a.1) Plasmonic nanoparticles employed for a near-field enhancement at 750 nm when being placed into perovskite. (b) Moth-eye metasurface modified PSC device and (b.1) its ability to localize the electro-magnetic field <cit.>. (c) Implementation of gold plasmonic nanoparticles into a transport layer of perovskite-based LEDs and (c.1) FDTD stimulation of the electromagnetic field distributed around the Au nanoparticle <cit.>. (d) Nanograted perovskite metasurface of photodetector and (d.1) irradiance-dependent responsivity plot at $\lambda$ = 635 nm <cit.> | 1806.08917_12.jpg |
arxivcap_2108.09620_2 | Title: Mittag--Leffler stability of numerical solutions to time fractional ODEs | Caption: Eigenvalues distributions with different parameters $D$ (Left); numerical solutions with $a=20, \delta x=1/64, \alpha=0.95, h=0.001$ and $u_0=5\sin(2\pi x)$ computed by $\mathcal{L}1$ scheme (Right). | 2108.09620_2.jpg |
arxivcap_2202.01537_5 | Title: Bending Graphs: Hierarchical Shape Matching using Gated Optimal Transport | Caption: Our GOT module consists of Sinkhorn iteration and recurrent feature propagation (in middle) in the shape graph. Promoting the feature propagation of more confident nodes, lets more information flow into weak nodes, consequently improving graph matching. | 2202.01537_5.jpg |
arxivcap_1802.01329_4 | Title: Reshaping of a Janus ring | Caption: Reshaping twisted ring filaments leading to equilibrium planar solutions at (a) $m=1$, $q=5$ (b) $m=4$, $q=12$ and to non-planar conformations with (c) single ($m=3$, $q=5$) and (d) multiple ($m=3$, $q=20$) self-contacts. | 1802.01329_4.jpg |
arxivcap_2303.06773_7 | Title: Protecting Classical-Quantum Signals in Free Space Optical Channels | Caption: RCI obtained using the protocol (solid lines), and by direct transmission (dashed lines) over a fixed loss channel with a probability of erasure $p_\mathrm{e}$. The values $\epsilon=\epsilon_\mathrm{det}$, and $\eta=1$ are used. A maximally entangled TMSV is considered as $\ket{\psi_\mathrm{AB}}$. | 2303.06773_7.jpg |
arxivcap_2305.05360_3 | Title: Anisotropic hybridization probed by polarization dependent x-ray absorption spectroscopy in VI3 van der Waals Mott ferromagnet | Caption: Effects of hybridization on XNLD spectra. (a-c) Calculated orbital occupations of the $e_{g}$ (orange triangles), $a_{1g}$ (pink circles) and $e_g^{\prime }$ (green squares) states per V atom as function of the symmetry adapted potentials $V_{e_g}$, $V_{a_{1g}}$ and $V_{e_g^\prime}$. At each time one of the potential was varied by keeping the others constant to the default values in Table <ref>. Ochre and cyan areas denote excess electrons on $e_g^\prime$ and $a_{1g}$ states respectively. (d-f) Calculated XAS and XNLD spectra at 20 K and 20$^\circ$ grazing incidence for $\tau=0$ eV and for $V_{a_{1g}}=1.098$ eV, $V_{e_g^\prime}=1.102$ eV (d), $V_{a_{1g}}=1.102$ eV, $V_{e_g^\prime}=1.098$ eV (e) and $V_{a_{1g}}=V_{e_g^\prime}=1.1$ eV (e). The schematics in the insets illustrate the corresponding state configuration and electronic occupation. Long red and short orange arrows correspond to integer and fractional electron count on V $3d$ states. The XAS spectra calculated with horizontal and vertical polarization are shown in blue and orange, while the XNLD signal is green. | 2305.05360_3.jpg |
arxivcap_2012.05350_5 | Title: Automatic Diagnosis of Malaria from Thin Blood Smear Images using Deep Convolutional Neural Network with Multi-Resolution Feature Fusion | Caption: Comparison of performance of different forms Dilationnet architectures with the DeepFusionNet architecture: Performance of DilationNet gets better with increasing resolutions of images. Considerable increase of performance is achieved by making the feature fusion with DeepFusionNet. | 2012.05350_5.jpg |
arxivcap_1401.8257_8 | Title: Online Clustering of Bandits | Caption: Results on synthetic datasets. Each plot displays the behavior of the ratio of the current cumulative regret of the algorithm ("Alg") to the current cumulative regret of RAN, where where "Alg" is either "CLUB" or "LinUCB-IND" or "LinUCB-ONE" or "GOBLIN"or "CLAIRVOYANT". The cluster sizes are balanced ($z=0$). From left to right, payoff noise steps from $0.1$ to $0.3$, and from top to bottom the number of clusters jumps from $2$ to $10$. | 1401.8257_8.jpg |
arxivcap_1912.05125_5 | Title: Bayesian Framework for Simultaneous Registration and Estimation of Noisy, Sparse and Fragmented Functional Data | Caption: (a) Simulated training data. (b) Fragmented simulated function with observations shown as black points. Posterior draws (transparent), pointwise mean (solid black) and $95\%$ credible interval (dashed) for the (c) amplitude and (d) phase components. (e) Composition of amplitude and phase, and PACE (red) and WPACE (blue) estimates. (f) Boxplots of $\mathbb{L}^2$ distances between the true function and estimated function. Each boxplot corresponds to a different approach: (1) plug-in, (2) MAP, and (3) pointwise estimators based on posterior samples from the proposed Bayesian model, and (4) PACE and (5) WPACE. | 1912.05125_5.jpg |
arxivcap_2206.12042_3 | Title: Experimental Demonstration of Quantum Pseudotelepathy | Caption: Experimental setup to play quantum magic square game. (a) The generation of two-photon Polarization- and OAM hyperentangled state. (b) Polarization interferometric prism (PIP). Implement a Polarization-OAM Controlled-NOT (CNOT) gate for single photons with the polarization acting as the control qubit and the OAM as the target qubit. The HWP with a WiFi sign represents HWP that changes angle according to QRN (quantum random number). (c) Polarization measurement (d) Dual-channel OAM readout by coherently convert OAM to the polarization by a swap gate (inset). (e) Quantum random number sends $0$, $1$ and $2$ to Alice and Bob, where $0$, $1$ and $2$ denote the first, second and third row or column in the game, respectively. OAM as the target qubit. | 2206.12042_3.jpg |
arxivcap_2010.12141_6 | Title: AdaCrowd: Unlabeled Scene Adaptation for Crowd Counting | Caption: We plot the variance for every feature $\phi$ extracted from guiding network for the last GBN layer (with dimension 128) in the decoder on the WorldExpo'10 dataset. (i) { blue} bars indicate the variance calculated over 10 images from the "same" scene; (ii) { green} bars indicate the variance calculated over 10 images where each is sampled from a "different" scene. In general, the variance for the features learned in the guiding network from the same scene is lower than that of images from different scenes. | 2010.12141_6.jpg |
arxivcap_1902.03928_1 | Title: The Degree of Fine-Tuning in our Universe -- and Others | Caption: Schematic representation of a small portion of the multiverse (adapted from <cit.>). Each individual universe within the larger ensemble is represented here as a separate expanding bubble. Every such region could in principle have a different realization of the laws of physics and/or different values for the cosmological parameters. The observable portion of our universe, depicted here as a white disk, is a small fraction of our entire universe - the region of space-time that has the same version of the laws of physics. The manner in which the various components of the multiverse are connected is not known, so that this depiction is heuristic. The number of theoretically expected universes is vastly larger than the number shown here. | 1902.03928_1.jpg |
arxivcap_1503.07328_1 | Title: Spectral weight redistribution in (LaNiO3)n/(LaMnO3)2 superlattices from optical spectroscopy | Caption: Reflectivity of the LNO/LMO superlattices. a) Room temperature reflectivities of the $n$=2,3,4,5 samples together with the reflectivity of pure LNO and that of the STO substrate. b) Temperature dependence of the reflectivity of the $n$=3 compound. c-e) Same for the $n$=4,5 and for pure LNO. Dashed lines correspond to intermediate temperatures 100, 150, 200, 250, and 350 K. In the inset of panel c) the data at 20 K and the relative fit are reported in the whole range. | 1503.07328_1.jpg |
arxivcap_2212.13102_6 | Title: Piezoelectric effect and diode effect in anapole and monopole superconductors | Caption: The Bogoliubov spectrum of the $A_g + i B_{3u}$ state with $\alpha_x = - \alpha_y = 0.4$, $\beta = 0$ (geometric effect model), and $T = 0.01$. We draw the spectrum on $k_z = 0$ and $k_y = 0, \pi/8, \pi/4, 3\pi/8$. | 2212.13102_6.jpg |
arxivcap_1211.6964_4 | Title: On the Bauschinger effect in supercoooled melts under shear: results from mode coupling theory and molecular dynamics simulations | Caption: Velocity profiles $v_x(y)$ as a function of position along the gradient direction $y$ for various times after application of a new flow direction. (a) Starting from a quiescent configuration (EQ), shear in $+x$ direction. (b) Starting from the resulting steady state (S) at $\gamma^s_w=0.4$, shear in $-x$ direction. Insets mark the points for which profiles are shown along the stress-strain curves. A total $100$ independent configurations are averaged within a small time window to obtain the depicted graphics. | 1211.6964_4.jpg |
arxivcap_1111.0053_2 | Title: Exploiting Subgraph Structure in Multi-Robot Path Planning | Caption: Examples of the four different subgraph structures. | 1111.0053_2.jpg |
arxivcap_2210.09459_5 | Title: Extensible Proxy for Efficient NAS | Caption: The loss surfaces of regression task with/without the barrier. | 2210.09459_5.jpg |
arxivcap_1906.05253_5 | Title: Search on the Replay Buffer: Bridging Planning and Reinforcement Learning | Caption: Better Distance Estimates: Without distributional RL, our method performs poorly. Ensembles contribute to a moderate increase in success rate, especially for distant goals. | 1906.05253_5.jpg |
arxivcap_2005.06188_7 | Title: Star clusters near and far; tracing star formation across cosmic time | Caption: Figure from <cit.>. The top panel shows the relation between the CO line width $\sigma$ and the gas surface density $\Sigma$ at a common resolution of 120pc for the discs of a sample of 15 nearby galaxies. The bottom left panel presents the mass-weighted distribution of the virial parameter $\alpha_{\rm vir}$ and the bottom right panel the distribution of turbulent pressure $P_{\rm turb}$ for the disc (circles) and center (star symbols) of all galaxies. The spread in molecular gas dynamical state and internal turbulent pressure is clearly visible within and between galaxies, in particular when comparing normal star-forming disc galaxies with a merger system such as the Antennae, or more quiescent galaxies such as M31. | 2005.06188_7.jpg |
arxivcap_2202.00194_3 | Title: The binding of atomic hydrogen on graphene from density functional theory and diffusion Monte Carlo calculations | Caption: Change of the electron density due to the adsorption of the H atom to the distorted graphene sheet (Eq. <ref>). $\rho_b$ from PBE calculations is shown from an oblique angle (A) and aligned along the $c$ axis (B). $\rho_b$ from DMC calculations (C) and (D) is shown from the same perspectives. Gold and blue represent a gain and loss of electron density, respectively. Note that there is a region of increased charge density at the C-H bond that is enveloped by a region of loss in the charge density. The binding density was visualized using an isovalue of 2.8$\times10^{-5}$ for DMC and 3.9$\times10^{-5}$ for PBE, in both cases capturing 95% of the differential charge density. | 2202.00194_3.jpg |
arxivcap_1806.08895_3 | Title: MRAttractor: Detecting Communities from Large-Scale Graphs | Caption: Original graph $G(V,E)$ where $|V|=12$, $|E|=16$. Each edge is associated with Jaccard distance in Equation <ref>. This graph is partitioned into four subgraph $G_{012}$, $G_{013}$, $G_{023}$ and $G_{123}$. The hash function $P(u)=u \mod p$ where $p=4$. The solid lines are main edges and the dot lines are rear edges. | 1806.08895_3.jpg |
arxivcap_2303.06773_8 | Title: Protecting Classical-Quantum Signals in Free Space Optical Channels | Caption: The protocol to transmit quantum and classical information (a). The protocol protects a single-mode quantum state against erasures by combining it with an entangled bipartite state, $\ket{\psi_\mathrm{AB}}$. Classical information is encoded using the displacement operation applied in mode 2. Bob monitors the channels via the LOs. The syndrome measurement corresponds to a dual homodyne measurement. A correction is applied to mode $1'$ to recover the original quantum state using the syndrome result. Dashed lines represent the transmission of classical information required to optimize the protocol. Turbulence simulations were obtained via phase screen simulations (b). Here, $r_t$ and $r_d$ represent aperture radii, and $L$ is the distance between transceivers. | 2303.06773_8.jpg |
arxivcap_1109.6664_5 | Title: Conditional pair distributions in many-body systems: Exact results for Poisson ensembles | Caption: (Color online) Evaluations of $g_{vv'}(\omega)$ at dimensions $D=1,2,3,\infty$ for $(nv,nv')=(0.2,0.2)$ (dotted lines) and $(nv,nv')=(1,0.5)$ (solid lines). | 1109.6664_5.jpg |
arxivcap_2004.03425_11 | Title: Elemental abundances in M31: [Fe/H] and [{\alpha}/Fe] in M31 Dwarf Galaxies Using Coadded Spectra | Caption: Top row: abundance comparison for all measurements for individual stars in our sample of MW GCs between the pipeline rewritten in Python <cit.> used for this study (W19), and that used for previous studies, written in IDL <cit.>. The left column shows and the right column shows for stars that pass the membership and quality criteria described in Section <ref>. Mean uncertainties are indicated in black for and in the left and right panels, respectively. Bottom row: the difference between the two pipelines as a function of S/N. We confirm that the two pipelines generally return similar results within the uncertainties, and the dispersion of the differences decreases as a function of increasing S/N. | 2004.03425_11.jpg |
arxivcap_2010.12141_7 | Title: AdaCrowd: Unlabeled Scene Adaptation for Crowd Counting | Caption: Performance vs. number of training scenes - we present the ablation study on the relation between network performance and number of training scenes on the WorldExpo'10 dataset with CSRNet and ResNet SFCN networks. | 2010.12141_7.jpg |
arxivcap_1709.03144_30 | Title: The 12C(a,g)16O reaction and its implications for stellar helium burning | Caption: (Color online) Comparison of all $E1$ and $E2$ cross sections measured to date. The early works of {PhysRevC.2.63, PhysRevC.2.2452, springerlink:10.1007/BF01415851} give only total cross sections. These are demarcated by an (*). As a standard for comparison, the $R$-matrix fit described later in this work is also shown. The solid red line shows the $E1$ contribution (except where only the total is given) while the red dashed line gives the $E2$ contribution. No normalization factors have been applied to the data. It should be noted that the region of astrophysical interest is at roughly $E_\text{c.m.} \approx$ 300 keV, far below the lowest energy measurements at $E_\text{c.m.} \approx$ 1 MeV. | 1709.03144_30.jpg |
arxivcap_2206.01897_4 | Title: Modeling of Textures to Predict Immune Cell Status and Survival of Brain Tumour Patients | Caption: Deep radiomic workflow. 1) Four MRI sequences (T1-WI, T2-WI, T1-CE and FLAIR) are acquired for GBM and LGG patients; ROI is then delineated using the 3D Slicer tool. 2) A group of parametric features (i.e., average, variance and weight) quantify the GMM curves that fit the histogram of 3D CNN feature maps. 3) Aggregate the deep radiomic features with clinical variable and immune cell markers for analysis using the RF models, Kaplan-Meier estimator and log-rank test. | 2206.01897_4.jpg |
arxivcap_2012.05350_6 | Title: Automatic Diagnosis of Malaria from Thin Blood Smear Images using Deep Convolutional Neural Network with Multi-Resolution Feature Fusion | Caption: Change of maximum obtained value in different evaluation metrics with increasing number of networks in DeepFusionNet: Performance of the DeepFusionNet improves with increasing number of networks for increasing the variations in the feature fusion scheme. | 2012.05350_6.jpg |
arxivcap_1806.08643_1 | Title: Diffusiophoretically induced interactions between chemically active and inert particles | Caption: Two spheres, one catalytically active ($S_1$) and the other catalytically inactive ($S_2$), are shown. The $S_1$ sphere, as a source of concentration gradients, converts species $A$ (reactant) to $B$ (product) in the reaction, $A + S_1 \rightarrow B + S_1$, which generates inhomogeneous concentration fields around the $S_2$ sphere. The $S_2$ sphere moves by the diffusiophoretic mechanisms due to the asymmetry of the concentration field in its vicinity. The numbers in the color bar indicate the normalized concentration of products ($B$). (The figure is constructed from simulation data described in the text. The sphere separation distance is $L/\sigma=3.5$.) | 1806.08643_1.jpg |
arxivcap_1802.01331_1 | Title: Relaminarization by steady modification of the streamwise velocity profile in a pipe | Caption: Sketch of the experimental facility used to control turbulent pipe flow with two different kinds of devices. Stereoscopic PIV measurements of the flow field are made downstream of the devices (see figure <ref> and <ref> for details). The flow direction is from left to right. Drawing not to scale. | 1802.01331_1.jpg |
arxivcap_2212.13102_9 | Title: Piezoelectric effect and diode effect in anapole and monopole superconductors | Caption: Results of the group velocity model: temperature dependence of (a) $q_0$, (b) $d^{\textrm{(SC)}}_{xxx}$ in the $A_g + i B_{3u}$ anapole state, and (c) $d^{\textrm{(SC)}}_{yzx}$ in the $A_g + i A_{u}$ monopole state. We set $\alpha_y = 0$, $\alpha (=\alpha_x) = 0.4$ and $\beta = 1$ or $0.7$. | 2212.13102_9.jpg |
arxivcap_2303.14662_8 | Title: OTAvatar: One-shot Talking Face Avatar with Controllable Tri-plane Rendering | Caption: 3D Consistency on multi-view dataset. We demonstrate additional visualization comparison on the Multiface dataset <cit.>, with more drastic camera view variation. All methods use the source frame to extract the identity feature, then extract 3DMM coefficients of pose and expression from the driving frame to generate the talking face. This subject is not included in the training set of any methods. | 2303.14662_8.jpg |
arxivcap_2206.11993_1 | Title: A Disability Lens towards Biases in GPT-3 Generated Open-Ended Languages | Caption: Average Toxicity score of GPT-3 generated text generated for three sentence prompts. X-axis represents the sentence prompts and Y-axis represents toxicity level ranging between 0 and 1. A higher value in the Y-axis indicates more toxicity in the sentence. | 2206.11993_1.jpg |
arxivcap_2202.01537_7 | Title: Bending Graphs: Hierarchical Shape Matching using Gated Optimal Transport | Caption: Results trained on SURREAL <cit.> and tested on Tosca <cit.>. In this experiment, we show the generalizability of the pipeline when trained on human shapes and tested on animal shapes. a) target shape. b) coarse matches c) source shape with the predicted correspondences d) ground-truth correspondences, e) normalized L2 error map. | 2202.01537_7.jpg |
arxivcap_2305.05360_4 | Title: Anisotropic hybridization probed by polarization dependent x-ray absorption spectroscopy in VI3 van der Waals Mott ferromagnet | Caption: XMCD on VI$_3$. Experimental XAS and corresponding XMCD spectra of VI$_3$ across the V $L_{2,3}$ edges measured at 1 T magnetic field and in normal incidence (NI): (a) FM phase (20 K) and (b) PM phase (65 K). (c) Theoretical XAS and XMCD spectra calculated at 20 K. Experimental and theoretical data are reported with dotted and solid lines respectively. Left- and right-circular polarization XAS and XNLD signals are shown respectively with blue, orange and green colors. | 2305.05360_4.jpg |
arxivcap_1503.07328_2 | Title: Spectral weight redistribution in (LaNiO3)n/(LaMnO3)2 superlattices from optical spectroscopy | Caption: Optical conductivity of the LNO/LMO superlattices as extracted from a Lorentz-Drude fitting. a) Room temperature optical conductivities of the $n$=2,3,4,5 samples and pure LNO and LMO. b) Temperature dependence of the optical conductivity of the $n$=3 compound. c-e) Same for the $n$=4,5 and for pure LNO. | 1503.07328_2.jpg |
arxivcap_1308.0826_1 | Title: Elemental Topological Insulator with a Tunable Fermi Level: Strained
\alpha-Sn on InSb(001) | Caption: Band schematics of for different situations without and with SO coupling (SOC), and with additional strain (double group representations refer to the diamond structure only, but are also used as band labels). Parity inversion exists independent of the presence of SOC as an inherent band feature. The TSS is hosted between second valence band ($\Gamma_7^-$) and conduction band, while the deep SO split-off band ($\Gamma_7^+$) is not involved. Strain opens a small band gap at the $\Gamma$-point. | 1308.0826_1.jpg |
arxivcap_1906.05253_6 | Title: Search on the Replay Buffer: Bridging Planning and Reinforcement Learning | Caption: Does SoRB Generalize? After training on 100 SUNCG houses, we collect random data in held-out houses to use for search in those new environments. Whether using depth images or RGB images, SoRB generalizes well to new houses, reaching almost 80% of goals 10 steps away, while goal-conditioned RL reaches less than 20% of these goals. Transparent lines correspond to average success rate across 22 held-out houses for each of three random seeds. | 1906.05253_6.jpg |
arxivcap_1902.03928_2 | Title: The Degree of Fine-Tuning in our Universe -- and Others | Caption: Schematic representation of the landscape of possible vacuum states of the universe (highly simplified). Each local minimum of the function shown here depicts a possible vacuum state for the universe, and hence a different possible realization of the fundamental parameters that determine the laws of physics. The number of vacua is expected to be enormously large <cit.>, with typical estimates of order $N_{\scriptscriptstyle V}\sim10^{500}$, or larger <cit.>, so that only a small portion of the entire landscape is shown here. | 1902.03928_2.jpg |
arxivcap_2202.13752_3 | Title: Discrete unified gas kinetic scheme for the conservative Allen-Cahn equation | Caption: The phase-field contour ($\phi=0$) of diagonal translation of a circular interface obtained by (a) DUGKS-AC, (b) DUGKS-I and (c) DUGKS-II with WENO-Z5 at Pe=500. Black solid line denotes the initial profile and red dashed line represents the reconstructed interface at $t=10T$. | 2202.13752_3.jpg |
arxivcap_2005.06188_8 | Title: Star clusters near and far; tracing star formation across cosmic time | Caption: {Left}: observed cluster formation efficiency,$\Gamma$, vs. star formation rate density, $\Sigma_{\rm SFR}$, relation obtained by combining data available in the literature. This plot contains only $\Gamma$ values determined with YSCs between 1-10 Myr. The data used are published by <cit.>, HiPEEC, Adamo et al (2020a, to be subm.). The solid line reproduces the fiducial model by <cit.> using the Kennicutt-Schmidt law, the dotted and the dashed lines reproduce the same model but applying the <cit.> conversion from $\Sigma_{\rm gas}$ to $\Sigma_{\rm SFR}$. The suggested constant $\Gamma$ (24 and 4.6 % estimated for the age range 1-10 and 10-100 Myr) values by <cit.> are included as horizontal dot-dashed lines. {Right}: Same as the left plot but for different age ranges, including clusters with ages $>$10 Myr (see text for details). We only include $\Gamma$ available in the literature that are estimated over a longer age interval. For details on the used cluster catalogues see <cit.> for the M31 data, and the LEGUS sample by Adamo et al (2020b, to be subm.). In both plots, empty symbols are used for plotting $\Gamma$ values detected on sub-regions of galaxies. See main text for discussion. | 2005.06188_8.jpg |
arxivcap_2212.13102_10 | Title: Piezoelectric effect and diode effect in anapole and monopole superconductors | Caption: Results of the geometric effect model: $\alpha$ dependence of (a) $q_0$, (b) $d^{\textrm{(SC)}}_{xxx}$ in the $A_g + i B_{3u}$ anapole state, and (c) $d^{\textrm{(SC)}}_{yzx}$ in the $A_g + i A_{u}$ monopole state. We set $\alpha = \alpha_x = -\alpha_y$, $\beta = 0$ and $T = 0.01$. | 2212.13102_10.jpg |
arxivcap_2010.12141_9 | Title: AdaCrowd: Unlabeled Scene Adaptation for Crowd Counting | Caption: We provide an overview of some failure cases caused by drastic changes in the target scene images due to illumination, occlusion or image quality. In the figure, (i) is a scene from the CityUHK-X dataset, while (ii)-(iv) are scenes from the PETS dataset. Nevertheless, our approach still performs better than alternative methods. | 2010.12141_9.jpg |
arxivcap_2202.00263_1 | Title: Fully Online Meta-Learning Without Task Boundaries | Caption: {Comparison of standard online meta-learning and FOML: In standard online meta-learning (e.g., FTML <cit.>), shown on the left, adaptation is performed on one task a time, and the algorithm "resets" the adaptation process at task boundaries. For example, a MAML-based method would reset the current parameters back to the meta-trained parameters. In our approach (right), knowledge of task boundaries is not required, and the algorithm continually keeps track of online parameters $\phi$ and meta-parameters $\theta$. The online parameters are simply updated on the latest data, and the meta-parameters are updated to "pull" the online parameters toward fast-adapting solutions via a MAML-style meta-update.} | 2202.00263_1.jpg |
arxivcap_2206.01867_1 | Title: SPGNet: Spatial Projection Guided 3D Human Pose Estimation in Low Dimensional Space | Caption: General workflow of our SPGNet, take 2D key points as input, and estimate the 3D pose with its global position. The 2D re-projection combines both estimations and loss to optimize the model. | 2206.01867_1.jpg |
arxivcap_1912.05125_9 | Title: Bayesian Framework for Simultaneous Registration and Estimation of Noisy, Sparse and Fragmented Functional Data | Caption: Posterior draws (transparent), pointwise mean (solid black) and $95\%$ credible interval (dashed) for the (c) amplitude and (d) phase components. (e) Composition of amplitude and phase, and PACE (red) and WPACE (blue) estimates. | 1912.05125_9.jpg |
arxivcap_2303.06733_2 | Title: Probing the Origin of Changing-look Quasar Transitions with Chandra | Caption: The dusty torus size, inferred from the rest-frame IR lag, vs. the rest-frame optical quasar luminosity. The black diamonds and gray crosses are lag measurements between $K$ band and optical band for nearby AGN <cit.>. The black open circles and gray solid dots are lag measurements between WISE $W1$ band and optical band for PG quasars <cit.> and SDSS quasars <cit.>. The orange solid diamonds are lag measurements in <cit.> for CLQs in literature. The blue (open), green (solid), and cyan (solid) are the lag measurements for CLQs in this work for J0224, J0829, and J1338, respectively. The magenta line is the R−L relation in <cit.>. | 2303.06733_2.jpg |
arxivcap_1802.01331_4 | Title: Relaminarization by steady modification of the streamwise velocity profile in a pipe | Caption: A comparison of the measured uncontrolled turbulent flow (ref) with the the law of the wall (fit from <cit.>, $w^{+}=2.5 \ln y^{+}+5.5$) and DNS in a pipe from <cit.>. The viscous sublayer is indicated by $w^{+}=y^{+}$. | 1802.01331_4.jpg |
arxivcap_2108.09610_1 | Title: The dwarf galaxy population in nearby clusters from the KIWICS survey | Caption: Sky Map of the full KIWICS sample. Red symbols correspond to the clusters analysed in this work. | 2108.09610_1.jpg |
arxivcap_2305.05260_1 | Title: Guided Focal Stack Refinement Network for Light Field Salient Object Detection | Caption: Three typical ways of focal stack handling, and the proposed model belongs to (c). | 2305.05260_1.jpg |
arxivcap_1104.3722_8 | Title: Investigating the Distribution of Password Choices | Caption: Number of users whose passwords are recovered by using the password ordering from one distribution to guess another. | 1104.3722_8.jpg |
arxivcap_2206.12076_10 | Title: Synthesizing Rolling Bearing Fault Samples in New Conditions: A framework based on a modified CGAN | Caption: Comparing the effect of each augmentation framework on the classifier performance when inner class is augmented - $t$ stands for true labeled samples. | 2206.12076_10.jpg |
arxivcap_1109.6664_6 | Title: Conditional pair distributions in many-body systems: Exact results for Poisson ensembles | Caption: (Color online) Contributions $g^{(iii)}_{vv'}(\omega)$ and $g^{(iv)}(\omega)$ (dashed and solid lines, respectively) to $g_{vv'}(\omega)$ at $v_+<\omega<\omega_{vv'}$, for $D=1,2,3,\infty$, and for the two ($v,v'$) conditions shown in Fig. <ref>. | 1109.6664_6.jpg |
arxivcap_1906.05303_2 | Title: Calculations for deep inelastic scattering using fast interpolation grid techniques at NNLO in QCD and the extraction of $\alpha_s$ from HERA data | Caption: Summary of values in comparison with the world average value. The inner error bars indicate experimental uncertainties, and the full errors the total uncertainty, comprised of the experimental and theoretical uncertainties. The lower set of values represent fits to data restricted to $\tilde\mu>28\,\GeV$. | 1906.05303_2.jpg |
arxivcap_1308.0826_2 | Title: Elemental Topological Insulator with a Tunable Fermi Level: Strained
\alpha-Sn on InSb(001) | Caption: a) GW quasi-particle calculation for bulk with realistic strain for InSb substrate (0.14% compressive). A small gap ($\sim$30 meV) opens at the $\Gamma$-point. b) Slab-layer LDA+U calculation for the (001) surface (52 layers, H-terminated) that shows the TSS with linear dispersion. Opposite spin character is coded in red and blue, respectively, reflecting a counter-clockwise spin orientation around the Dirac cone. c) Close-up view of the TSS bands, based on a slab calculation performed in 256 layers using a tight-binding parametrization of a bulk GW quasi-particle calculation (see Supp. Mat. [25]). The TSS is clearly seen to emerge from the second valence band (colors: spin, symbol size: surface character within topmost 30 layers), with a Dirac point $\sim$ 0.1 eV below . | 1308.0826_2.jpg |
arxivcap_2210.09459_6 | Title: Extensible Proxy for Efficient NAS | Caption: Visualize the architecture found by Eproxy and Eproxy+DPS on ImageNet-DARTS search space. | 2210.09459_6.jpg |
arxivcap_2212.13102_11 | Title: Piezoelectric effect and diode effect in anapole and monopole superconductors | Caption: Results of the geometric effect model: temperature dependence of (a) $q_0$, (b) $d^{\textrm{(SC)}}_{xxx}$ in the $A_g + i B_{3u}$ anapole state, and (c) $d^{\textrm{(SC)}}_{yzx}$ in the $A_g + i A_{u}$ monopole state. We set $\alpha = \alpha_x = -\alpha_y = 0.4$ and $\beta = 0$. | 2212.13102_11.jpg |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.