dedup-isc-ft-v107-score
float64 0.32
1
| uid
stringlengths 32
32
| text
stringlengths 0
32.5k
| paper_id
stringlengths 1
14
| original_image_filename
stringlengths 5
222
|
|---|---|---|---|---|
0.464011 | ec5c0867b22c45c6a96764f8ef44a994 | Crystal strucure of NdAlSi.(A) Structure of the non-centrosymmetric NdAlSi with space group I4_1md (109) and its associated first Brillouin zone shown in (B). There are two simple mirror planes, m_x and m_y, and two glide mirrors, m_xy and m_xy̅. These are shown in (B) as the light red and green planes, with the Z–Γ–X path lying in the green m_xy mirror plane and the Γ–S–N–S^'–Z–Γ path lying in the red m_x mirror plane. (C) Refinement of single crystal nuclear neutron diffraction data collected at 100 K. The Second Harmonic Generation signal is shown as an inset. (D) Temperature dependence of the magnetic heat capacity and magnetic entropy of NdAlSi, which were isolated from the net heat capacity by subtracting the heat capacity of non-magnetic LaAlSi. | 2012.12970 | Fig1.jpg |
0.438748 | dfb10a15db4b4d6c828db372905a5a56 | Band structures, Fermi surfaces, and Weyl node positions as calculated within PBE+U+SOC. (A-B) Weyl nodes W_1, (C-D) Weyl nodes W_2. (E-F) Weyl nodes W'_2. (G-H) Weyl nodes W_3. (I-J) Weyl nodes W'_3. (K-L) Weyl nodes W”_3. | 2012.12970 | Fig2-SI.jpg |
0.404907 | ef918506142247d08a0472a2ecc94167 | Magnetic phase transitions in NdAlSi. (A) temperature dependence of the inverse susceptibility for field applied along the c axis (χ_c) as well as its ratio with the in-plane susceptibility (χ_a). (B) is the ab plane longitudinal electrical resistivity of NdAlSi versus temperature. (C) and (D) respectively show the temperature dependence of the ordered moment and the associated wavevectors. | 2012.12970 | Fig2.jpg |
0.409602 | 547676dab4324c649dd4ef1a9fd6af32 | Commensurate ferrimagnetic spin structure of NdAlSi. (A) Rocking scans at 𝐐=(101) and 𝐐=(004) whose magnetic intensities are consistent with the FM spin structure shown in the left inset. (B) shows the L dependence of the 𝐐=(2/32/3L) Bragg peaks at T=1.6 K, which have been normalized to the (2/32/30) Bragg peaks, and multiplied by the ratio of their form factors. The calculated L dependence for various anisotropies are plotted and reveals the AFM 𝐤_ com=(2/32/30) component is Ising like (see bottom inset). The red spins were refined to 3.8(4)μ_B, the blue spins to half this size. (C) 3D representation of the NdAlSi spin structure, which is obtained by adding the structures shown in the insets of (A) and (B). (D) shows the effective 1D spin configuration of (C) projected to z = 0 and plotted along the [110] direction. (E) shows the spin structure with the c axis coming out of the page. The c axis field dependence of the magnetization as well as the AFM 𝐤_ com component at T = 1.6 K are shown in (F) and the electrical resistivity in (G) for various temperatures. | 2012.12970 | Fig3.jpg |
0.476337 | aa0c21f2fe86464a83e7ab30aff8771b | Quantum oscillations in NdAlSi. (A) Quantum oscillations (QO) recorded above and below the transition temperature T_inc(H=0) (separated by the horizontal dashed line), above and below the metamagnetic (MM) transition field H_c (separated by the slanted line). The FFT spectra of the oscillations for the PM phase, the AFM (2/32/30) phase, and the FM phase are respectively shown in B, C and D. (E) The temperature evolution of the magnetic incommensurability q_m (see Fig. <ref>), and quantum oscillation frequencies k_α and k_β in different states. Note that the frequencies are shown in ^-1. | 2012.12970 | Fig4.jpg |
0.449086 | 9603a53d39bb42c19540af1fcd3a9c61 | Electronic band structure for the ferromagnetic phase of NdAlSi. (A) Ferromagnetic PBE+U+SOC band structure. (B-C) Ferromagnetic PBE+SOC Weyl nodes in top and side views respectively. The six types of Weyl nodes are marked as W_1, W_2, W'_2, W_3, W'_3 and W”_3. (D) and (E) are respectively the top and side view of the ferromagnetic Fermi surface whose Fermi energy was determined by the quantum oscillations analysis. The red color represents the valence band whereas the blue color is the conduction band. | 2012.12970 | Fig5.jpg |
0.43888 | 5e34b78b528d499eb60493d361df03cf | Magnetic structure refinement of NdAlSi. Real part of the basis vectors for a 𝐤=(000) and 𝐤=(2/32/30) spin structure within space group 109 are shown in A and B respectively. Rocking scans at the 𝐐=(2/32/30) and 𝐐=(1/31/30) Bragg positions are compared in C. Note that θ for the 𝐐=(1/31/30) Bragg peak was translated by 4.95. The observed magnetic structure factor as a function of the calculated structure factor for both the FM 𝐤=(000) and the AFM 𝐤=(2/32/30) components are plotted in D. | 2012.12970 | FigSNeutron.jpg |
0.486325 | a1437313c16145acb65a57ef127584aa | High-Field Quantum Oscillations in NdAlSi. Resistivity measured up to 35 T at various temperatures with field along the c-axis in A. Shubnikov–de Haas (SdH) oscillations appear and grow as the field increases. SdH oscillations as a function of 1/H at different temperatures in B. The change in the oscillatory frequencies can be seen from the shift of peak positions as the temperature increases. FFT spectrum based on the oscillations in panel B is reported in C. Three distinct Fermi pockets are identified: δ, β, and γ, and their frequencies at 0.316 K are 40 T, 66 T, and 101 T respectively. The peaks marked by 2β (F_2β=135 T at 0.316 K) are identified as the second harmonics of β. Note that δ frequency is different from α frequency (F_α = 20 T at T=1.8 K) in the d-u-u phase. The effective masses extracted for each Fermi pocket using a standard Lifshitz-Kosevich formula are reported in D. | 2012.12970 | NdAlSi_QO_SI.jpg |
0.506189 | 1840c5b3343e43a39d236522951d3654 | T∩^2, 2T∩^2, and 3T∩^2 from Example <ref>, yielding the 2nd, 3rd, and 4th triangular numbers, respectively. | 2012.12976 | fig1.jpg |
0.505964 | bdd0a43674d54d65a5f19564ba87df98 | Taking s=4 and t=6 in Example <ref>, we examine the line segment of x,y≥ 0 such that 4x+6y=4· 6, which has endpoints (0,4) and (6,0). Therefore f(4,6)=3=(4,6)+1. | 2012.12976 | fig10.jpg |
0.476923 | fba0696768214f759bdcc0e2a56b6c96 | The integer hull (solid line) of P_5 (dotted line), in Example <ref> (cf., Figure <ref>). When t is odd, the integer hull of P_t is an octagon with vertices (0,±t-1/2),(±t-3/2,±t-1/2),(±t-1/2,0),(±t-1/2,∓t-3/2). When t is even, the vertices of P_t are integers, so the vertices of the integer hull are simply the vertices of P_t: (±t-2/2,±t/2) and (±t/2,∓t-2/2); see Example <ref>. | 2012.12976 | fig11.jpg |
0.449669 | 3a023048235249a196fdb9c90afd40e2 | Q∩^2, 2Q∩^2, and 3Q∩^2 from Example <ref>. | 2012.12976 | fig2.jpg |
0.446649 | 6fd824f8bc4745aa9af354cf1ecd88a2 | On the left is P from Example <ref>, the tetrahedron formed by the convex hull of the three labeled vertices and the origin; the f_P(1)=P∩^2=4 integer points are shown. On the right is 3P, and shaded is the cross-section of 3P at x=2; this cross-section has 6 integer points (the third triangular number), as shown. | 2012.12976 | fig3.jpg |
0.455095 | aeefa80b62074ba0ba739fd9e031180f | T∩^2, 3T∩^2, and an overlay of 3T, 4T, 5T, from Example <ref>. | 2012.12976 | fig4.jpg |
0.405026 | 49598f66324a442f894dfd65f74ab38b | The triangle 75P, from Example <ref>. The three vertices are labeled, as are the f_T(75)=5 integer points in 75P; for example, the point (1,1,3)∈ 75P shows that exactly 75 McNuggets can be purchased by ordering one box of 6, one box of 9, and three boxes of 20. | 2012.12976 | fig5.jpg |
0.372773 | d867f4f463a0410286edf59866d13013 | P_2,6∩^2 and P_1,7∩^2 from Example <ref>. | 2012.12976 | fig6.jpg |
0.429512 | c14d400513344af9b8b1e125897bf1c2 | The graph, G, in Example <ref>. | 2012.12976 | fig7.jpg |
0.466256 | 6d625c6d961d4929bf0056b8cbba30f4 | A 24-magic square; see Example <ref>. | 2012.12976 | fig7b.jpg |
0.473586 | 9b6d44691997417681ad99ea00731873 | Diagonal line segments are 3x+5y=n, for n=7,16,25, constrained by x,y≥ 0. We see that 7∉ S, because there are no integer points on the line segment. On the other hand 16∈ S and 25∈ S; in fact, the two integer points for n=25 correspond to two ways to write 25 using 3's and 5's: 25=0· 3+5· 5=5· 3+2· 5. Furthermore, we can see the second quantifier elimination step used in Example <ref>: if there are some x,y∈_≥ 0 with 3x+5y=n, then we may take y to be 0, 1, or 2; indeed, if (x,y) is on the line segment with y≥ 3, then (x+5,y-3) is also on the line segment, e.g., (0,5) and (5,2) are on the line segment corresponding to n=25. | 2012.12976 | fig8.jpg |
0.51107 | 3a4ca318ee4a4ea2acb455d5b7189af0 | The twisting square P_t in Example <ref>, for t=1,2,3,4,5,6. Even t are in solid lines and odd t in dotted lines. | 2012.12976 | fig9.jpg |
0.384032 | bdf69a412345400bb37eab4942b4f0c0 | Source parameter space with optimal colored for particular observation times drawn as green contour lines. We impose that the signal cannot spin-up by more than 1 Hz/s and cannot reach a frequency greater than 2048 Hz. The optimal choices for and do not behave linearly at high frequencies, emphasizing the need for optimizing the choice of these two parameters. Moreover it is clear that for lower frequencies and lower chirp masses, we should observe for longer with higher , which is consistent with the fact that signals in that portion of the parameter space are more similar to continuous waves than transient continuous waves. To create this plot, we use the advanced LIGO/Virgo sensitivity curves to determine S_n, and evaluate S for a range of and at each point f_0 and ℳ in the parameter space, and plot the combination of and that maximizes S. | 2012.12983 | best_fft_plots_linspace.jpg |
0.423206 | 96c0e4f953174db7bf303bfcd5af364b | We show the computational cost to perform the Generalized Frequency-Hough in a directed search as a function of the source parameter space. We assume a known sky position, so the only limit on the observation time or comes from the source's intrinsic properties (its lifetime, its spin-up, etc.). The computational cost varies greatly across this space, with more “transient” signals requiring a lot less time than more “continuous” ones. White space in the bottom left-hand corner appears because for those parameter points, the linear approximation to equation (<ref>), equation (<ref>), is valid. There are not any points in the right-hand corner because systems with these frequencies and chirp masses would either spin-up too quickly (ḟ>1 Hz/s) or merge too quickly (<1 s). We observe a maximum as a “yellow” curve because we fix the observation time to be at most one year, and at points below this curve, the Fast Fourier Transform time is increasing, thus resulting in a decrease in computational time. Based on our estimates, on 2000 Quad-Core Intel Core i7 cores in parallel, a directed search using the Generalized Frequency-Hough on one detector's data would take only a few days. | 2012.12983 | dir_comp_cost_optimal_FFT.jpg |
0.62071 | 4a645b6dd783451794389dde51973aba | PBH mass distribution normalized to f_ PBH = 1, for Case 2. This model includes the equation of state reduction effects at the Quantum Chromodynamics transition on PBH formation for primordial power spectra with indices n_ s = 0.95/0.96/0.97 (yellow, red and blue curves). | 2012.12983 | fPBHthermal.jpg |
0.410041 | 8a13273845234d86af874aea5afa779d | Expected limits on the dark matter fraction made of PBHs as a function of the chirp mass, for primordial binaries in Case 1 - agnostic mass function - (solid lines) and Case 2 - thermal mass function - (dashed lines), and for binaries formed through tidal capture in Case 2 (dotted-dashed lines). The different colors represent the limits from galactic binaries (gal), from the galactic center (GC) and in the solar system vicinity (sol. sys.), for the expected sensitivities of advanced LIGO/Virgo and ET. The dotted elliptic region represents constraints from the Optical Gravitational Lensing Experiment (OGLE) and the Subaru Hyper Suprime-Cam (HSC) <cit.> for comparison. | 2012.12983 | limits_newversion.jpg |
0.431602 | 21130e4cbdef4750904610613e905283 | Time before merger as a function of frequency and chirp mass. The widely distributed signal durations in the parameter space imply that different techniques are needed to probe the existence of PBHs at different masses. | 2012.12983 | mergtime.jpg |
0.407345 | 3f0f375574e84d9d98b4eb06a77f71f9 | PBH merging rates within a sphere of radius d centered on the solar system, for Case 1 with f_ PBH = 1. The colored solid lines represent LIGO/Virgo-O2 (red) and Einstein Telescope (ET) sensitivities at 50 Hz and 10 Hz, respectively, for the analyses methods described later in section <ref>. Dashed and dash-dotted lines correspond to one binary inspiral at f_ PBH = 1 and f_ PBH = 0.1, respectively, for LIGO/Virgo-O2 (red) at 50 Hz and ET (yellow) at 10 Hz. When the solid curves go above the dashed or dash-dotted lines, we can claim that one binary inspiral at a particular chirp mass and distance could be detected by LIGO/Virgo or Einstein Telescope. | 2012.12983 | solsys_case1b.jpg |
0.445492 | df8f99e1a8fa474dbf90330895e49ef1 | Spin-up as a function of chirp mass with orbital frequency colored. A green line representing the maximum spin-up to which continuous-wave searches have considered is also plotted <cit.>. Colored points below the green line, meaning smaller spin-ups, represent possible masses of inspiraling PBHs that can be probed with continuous-wave methods. The maximum chirp mass of a PBH merger that we could detect with continuous-wave methods is ∼5× 10^-4M_⊙. Transient continuous gravitational-wave methods are necessary to exhaustively constrain larger PBH chirp masses. | 2012.12983 | spinup_vs_mc_freq_colored.jpg |
0.479633 | 77f601c5e6e14eeeab8cdaf2b8fde627 | From cell cultures to quantitative biological information. a The standard workflow entails chemically staining the cell structures of interest, imaging them using fluorescence microscopy (in multiple light channels), and, finally, using these fluorescence images to retrieve quantitative biologically-relevant measures about the cell structures of interest. b The deep-learning-powered approach we propose replaces the chemical-staining and fluorescence microscopy with a conditional generative adversarial neural network (cGAN) that uses brightfield images to generate virtual fluorescence-stained images. | 2012.12986 | Fig1.jpg |
0.463344 | c9d069e6cd8b4b8ca5272d6134bae670 | Conditional generative adversarial neural network (cGAN) for virtual staining. The generator transforms an input stack of brightfield images into virtually-stained fluorescence images of lipid droplets, cytoplasm, and nuclei, using a U-Net architecture with the most condensed layer being replaced by two residual network (ResNet) blocks <cit.>. In the first layer of the generator, we normalize each input channel (i.e., each brighfield z-slice) in the range [-1,1] using Equation (<ref>). The U-Net encoder consists of convolutional blocks followed by max-pooling layers for downsampling. Each convolutional block contains two paths (a sequence of two 3 × 3 convolutional layers, and the identity operation), which are merged by concatenation. The U-Net decoder uses bilinear interpolations for upsampling followed by concatenations layers and convolutional blocks. Next, the hyperbolic tangent activation transforms the output to the range [-1, 1]. In the last layer of the U-Net, the network learns to denormalize the output images back to original pixel values by scaling and adding an offset to the output. Every layer in the generator, except the last two layers and the pooling layers, is followed by an instance normalization and a leaky ReLU activation. The discriminator is designed similar to the PatchGan discriminator <cit.> and receives both the brightfield images and fluorescence images (either the target fluorescence images or those predicted by the generator). The inputs to the discriminator are normalized as those to the generator. The discriminator convolutional blocks consist of 4 × 4 strided convolutions for downsampling. In all layers in the discriminator, we use instance normalization (with no learnable parameters) and leaky ReLU activation. Finally, the discriminator outputs a matrix containing the predicted probability for each patch of 32 × 32 pixels. | 2012.12986 | Fig2.jpg |
0.452593 | 55d3245b691a44189c2a5c930e9e3b16 | Virtually-stained fluorescence images. a Brightfield image and corresponding b-d chemically-stained and e-g virtually-stained fluorescence images for lipid droplets, cytoplasm and nuclei. h-o Enlarged crops corresponding to the dotted boxes in a-g. The lipid droplets are clearly visible in the brighfield image (a and h) thanks to their high refractive index so that the cGAN manages to accurately predict the chemically-stained images (b and i) generating accurate virtual stainings (e and m), even reproducing some details of the internal structure of the lipid droplets (indicated by the arrows in i and m). The chemical staining of the cytoplasm (c and j) is also closely reproduced by the virtual staining (f and n). The virtually-stained nuclei (g and o) deviate more prominently from the chemically-stained ones (d and k), especially in the details of both their shape and texture, which can be explained by the fact that the nuclei are not clearly visible in the brightfield image so that the cGAN seems to use the surrounding cell structures to infer the presence and properties of the nuclei shape. | 2012.12986 | Fig3.jpg |
0.382519 | b7051bc18e4a49679429ecb7af2e7985 | Quantitative information from chemically-stained and virtually-stained fluorescence images. Segmentation obtained using CellProfiler (https://cellprofiler.org, version 4.07 <cit.>) of a-c chemically-stained target image and d-f virtually-stained generated image for lipid-droplet, cytoplasm and nuclei. Probability distribution for g-i the size and j-m the mean intensity of the individual lipid droplets, cytoplasmatic regions, and nuclei identified by CellProfiler for the chemically-stained (gray histograms) and virtually-stained (colored histograms) segmentations. n-p Total cell structure count, mean area, integrated intensity, mean intensity, and standard deviation of the mean intensity for the lipid droplets, cytoplasmatic regions, and nuclei identified by CellProfiler in the virtually-stained segmentations (colored outlines) normalized to those identified in the chemically-stained segmentations (gray outlines). | 2012.12986 | Fig4.jpg |
0.406283 | 3eb9c7c24cc941eb965a2db696ad2533 | Data augmentation technique applied in movement path | 2012.12987 | dataaugmentation.jpg |
0.459366 | 40d37d0193bb440fa2fb4f367ac80e09 | Samples of the path data plotted as images after rescaling | 2012.12987 | dataimageplot.jpg |
0.425469 | 90306809c6a24ccdad7b9d731fdc3f56 | Direct, pacing and random | 2012.12987 | delaunay1.jpg |
0.411306 | 6a1e1617214e462b945358a5ed78d69d | Lapping | 2012.12987 | delaunay2.jpg |
0.41608 | 4fad775f2eeb4cc4a61617a40455f6cb | Developed application for data generation. The user interacts with the floor-plant to place movement points, define the hour of the movement and if it was wandering. On the right side of the page shows the statistics created to visualize the generated data. In the upper right a menu that shows all the registered movement dates and hour. | 2012.12987 | geradordados.jpg |
0.516379 | 3eef4192fce74e5bba08d4680d988cb6 | Accuracy per epoch of the model | 2012.12987 | graficotreinamento-accuracy.jpg |
0.425979 | 9c4fd901a845473b9d7bb73a8e380f4d | Learning curve representation of the model | 2012.12987 | graficotreinamento-loss.jpg |
0.429131 | a6b7a06040994cf189af38e2b63b6675 | Architecture of our neural network model. | 2012.12987 | modelo-redesneurais.jpg |
0.488415 | 5a80fb4b07ba47bd993436b637e97aca | Sample of a random path generated with wandering activity | 2012.12987 | randomprediction.jpg |
0.454901 | 338b1a3fd69446b8981923aaf9360fbc | In the flow of time. This original painting tells the story of the study, describing our journey as we delve into the hidden secrets behind large–format artworks. A fluctuating and changing process, our multi–analytical study's use of novel software tools and multispectral imaging allowed us to fantasize, even for a second, that we could stop the passage of time and capture the artist's creative process, both materially and conceptually. As the colours flow through our fingers, we perceive individual pigment crystals, sense the damage caused by the passage of time, understand the materials and layers that come together to form the painting, and almost hear the artist's voice as we unveil his intentions. Although created using contemporary techniques such as acrylic and collage on canvas, this piece was inspired by Carlo Ferrario's creative process. His colour palette influenced our own, as we selected pigments found in Musas I and Musas II such as ultramarine blue, vermilion, and white. Our painting is structured similarly to Ferrario's, with two layers of ground, two layers of paint, and a final layer of varnish. Art, like history and time, slips inexorably through our fingers. Today, we capture in this study an instant of that flow, gaining new perspectives and revealing some of the hidden secrets behind large-format paintings. | 2012.12989 | Figure1.jpg |
0.476207 | f5bf39bdc8b14b53959bbb3f891d4421 | Six views at once: Multispectral Imaging of Musas I (top panel) and Musas II (bottom panel), (a) Visible (VIS), (b) infrared (IR), (c) infrared false colour (IRFC), (d) ultraviolet fluorescence (UVF) and (e) ultraviolet reflectance (UVR) multispectral images of Musas I and Musas II and (f) sample locations on the grid. The M1-XXY codification of sample names indicates: M1 = Painting Musas I, M2 = Painting Musas II, XX = location on the grid, Y = visible colour of the sample (B = blue, R = red, P = pink, W = white). | 2012.12989 | Figure2.jpg |
0.423622 | 8009dfdf557c4480a043dc6b6a05ba0f | A closer look: Execution of RegionOfInterest software. (a) The program opens the image for general user inspection. (b) In the first stage, the user zooms in on a region of interest. (c) The user hand–draws a contour to select the area for analysis. (d) The program converts the selected pixels from HSV to Intensity space and automatically generates a "Counts vs Intensity" histogram indicating pixel intensity distribution ranging from 0 to 255. (e) In the second stage, the user chooses the range from the first histogram. (f) Pixels in the full image that fall within this intensity range are displayed in a new colour image. (g) A second intensity histogram is generated, comparing the intensity distribution of all pixels in the original image (blue line) with that of the selected pixels in the new image (orange line). Note that the y–axis shows different scales for the blue and orange curves, left and right, respectively. | 2012.12989 | Figure3.jpg |
0.40404 | 482e5399b6a04096858734ceb15d3b36 | Through the lens: Stratigraphy of a 120–year–old large–format painting. (a) Panoramic Optical Microscopy (OM) image of the cross–sectioned sample M1–75W, observed by reflected visible light at 5X. (b) Detail of the sample observed by MO at 20X showing the main layers: (1) Top paint layer, (2) Intermediate paint layer, (3) Second ground layer, (4) First ground layer. (c) Fluorescence Microscopy showing fluorescent particles in layer 4. (d) Scanning Electron Microscopy (SEM) showing the area analyzed by SEM–EDX using the mapping tool. (e) The main elements in each layer. | 2012.12989 | Figure4.jpg |
0.455405 | 2a363560682c47d7a982c72cf4df1f0e | Between particles: Measuring crystals with CrystalDistribution software (a) The user opens a microscope image and select the region containing crystals of interest. (b) The program applies masks or "cuts" identifying the crystals and saves the images. Each image shows the crystals of a specific colour. (c) The program calculates the crystals' area and diameter, and generates a histogram for each colour. | 2012.12989 | Figure5.jpg |
0.39412 | 159b293957664069a1a846ea57320d9e | In low frequency: Raman spectroscopy of the pigments. (a) Comparison of Raman spectra for reference pigments and those of the samples (b) Orangey–pink crystal from sample M1–54P corresponds to lead red, (c) Green and blue crystals from sample M2–69P correspond to viridian and ultramarine, respectively. (d) Red and yellow crystals from sample M1–45B correspond to vermilion and chrome yellow, respectively. | 2012.12989 | Figure6.jpg |
0.423748 | e582ee880dc24440a4e9dc55f50c26d4 | Traces of time: Categorising damage observed in the paintings. Visible photographs of (a) Musas II and (b) Musas I with the working grid. (a.6) Musas II and (b.6) Musas I diagrams showing locations of the ten most significant types of damage, with visible and UVF images of each: a.1.) Loss of paint and exposed canvas. a.2.) Spots. a.3.) Craquelure. a.4.) Fungi. a.5.) Holes. b.1.) Detachment of canvas (bubbles). b.2.) Moisture marks. b.3.) Cracks. b.4.) Human fingerprints. b.5.) Insect specks and perforation. | 2012.12989 | Figure7.jpg |
0.493641 | c2bed151ea354e1d9d79713e3e2d222b | Conservation in the tropics: Monitoring the paintings' environment. (a) National Theatre of Costa Rica in San José (NTCR), Costa Rica, Central America. (b) Musas I and Musas II are located on the ceiling, at a height of about 3.5 m in the ladiesâ lounge on the ground floor of the Theatre. (c) Location of the environmental sensors. (d) Temperature and (e) humidity variations in the paintings' microclimate. | 2012.12989 | Figure8.jpg |
0.421861 | 0617de29a5e14a85894b106294c3572d | Architectures of the Faster R-CNN (on the left) and Cascade R-CNN (on the right). F is the feature extractor, N_is are network bodies, R is the region proposals, L_is are label predictions and R_is are the bounding box predictions. | 2012.12991 | casc.jpg |
0.428087 | 142455e2aa144221996bf68f4b9e92d5 | Architecture of SyNet. N_is are network bodies containing convolutional layers, F is the feature extractor, CN is the CenterNet body, R is the region proposals, L_is are label predictions, R_is are bounding box predictions, E is the ensemble function and L_F, R_F are final label and bounding box predictions, respectively. | 2012.12991 | comb.jpg |
0.415043 | 946e80d8b300468db523b371f0278947 | A. bands with the highest correlation for each property B. NIRS spectra and bands with relative peak positions for soil costituents absorption | 2012.12995 | Fig5.jpg |
0.475059 | 01f63bae0f9846108c1c1ad74a1feb7e | Regression results: A. pH, B. Organic matter, C. Ca in log scale, D. Mg. For each property, we present the result with the best ML model (red). And the results simulate chemometric techniques such as the regression result with the band (blue) with the highest correlation and the PLSR (green). E. Table with models comparison for property: correlation ρ_test and determination R_test^2 coefficients, and MSE in the test set (30%). The models presented are the three result of the ML regression (background in red for the model with the best results), and the two approaches that simulate chemometric techniques (1) linear regression with the highest correlated band (background in blue) and (2) PLSR (background in green). Colors in the background correspond to the colors in Figures A-D | 2012.12995 | Fig_reg_tblv3_2.jpg |
0.45153 | 7cdd09bda58f46ffb2fb8d71615dc675 | Strategy for regression and classification model selection from vis-NIRS data. Both start from the same data and features. At left, the steps that are taken to obtain the best regression model for each property; and at right, the classification steps | 2012.12995 | Strategy.jpg |
0.452489 | 08d0600411454959808676aec5656b20 | A. Sampling area for 653 points. Sample area in Santander and Boyaca district C. and in the country B. | 2012.12995 | Study_area_paper_3.jpg |
0.443202 | 165e3a2ad47e495aa82b6baa8419aee5 | The flight paths while capturing the five Davos tiles. | 2012.12996 | Davos_5tiles_flight_path.jpg |
0.451362 | 39253c97104c42e98a7fd2302d6857ef | Our data processing and segmentation pipeline. | 2012.12996 | dataprocessingflowchart.jpg |
0.371135 | 830cdcf4409843e5affd53c042b7dc2d | The weighted accuracy (left) and average IoU (right) for the models and ensembles (bars) trained on the same or different cities. | 2012.12996 | three_test_tiles_ensembling_barchart.jpg |
0.419431 | e1fd965968fd4fd69428a28cce13790f | The weighted accuracy (left) and the average IoU (right) for different models and model ensembles (rows) evaluated on different cities (columns). We consider: three models trained on data from each city independently; one model trained on data from all three cities; three ensembles of pairs of single-city models; one ensemble of three single-city models. | 2012.12996 | three_test_tiles_weighted_accuracy_IoU_heatmaps.jpg |
0.438721 | 63b121d90cfa47e5ae0a263ef9b2bb3f | Upper panel: The best fitting RV curve derived from the orbital solution to the 28 phase-folded APOGEE RVs for the WDMS system 2M10243847+1624582. Two periods are shown for clarity. Lower panel: The residuals to the fit shown in the upper panel. | 2012.12997 | 2M10243847+1624582_phased_twice_MCMC.jpg |
0.43646 | 2c85ffeaed5542579a8f2144e797056d | The same as Figure <ref> but for the eleven APOGEE RV epochs for 2M10552625+4729228. The error bars are smaller than the plotted points in the upper panel. | 2012.12997 | 2M10552625+4729228_phased_twice_MCMC.jpg |
0.468877 | 30e4ca71962d498691dc92563b487581 | The same as Figure <ref> but for the 33 APOGEE RV epochs for 2M11463394+0055104. | 2012.12997 | 2M11463394+0055104_phased_twice_MCMC.jpg |
0.451757 | 0c26e510ff2f4e3c8e904624e3cb4422 | The same as Figure <ref> but for the six APOGEE RV epochs for 2M13054173+3037005 and assuming a circular orbit. While the number of epochs is small, their phase coverage is very good for this adopted solution, while the RV amplitude is very large, which results in a solution that converges to a tight match with the data. | 2012.12997 | 2M13054173+3037005min_cut_e0_phased_twice_MCMC.jpg |
0.414342 | 5859619fb16a423d8e055c7adb36c2c7 | The same as Figure <ref> but for the seven APOGEE RV epochs for 2M14544500+4626456. While the number of epochs is relatively small, the combination of very good phase coverage and a large RV amplitude means that a very good solution is possible, under the assumption of a circular orbit. | 2012.12997 | 2M14544500+4626456_e0_phased_twice_MCMC.jpg |
0.422167 | 829b43d79410415ab3228fe24abdbf57 | Mass ratio, q, versus period for PCE systems with previously derived orbital parameters and component masses (orange). The five systems with orbital parameters derived in this work are also shown (blue). The mass ratios for the latter represent lower limits on q because we derive the minimum WD mass from the Keplerian orbital parameters. | 2012.12997 | WDMS_orbits_update.jpg |
0.406358 | 06db2b7a554b4818970061552541d986 | Top: The metallicity distribution of the WB and PCE systems listed in Table <ref> shown in blue and orange, respectively. For comparison we also show in black the distribution of APOGEE MS stars sharing similar effective temperatures and surface gravities as the MS primaries in the WB and PCE samples. Bottom: Cumulative distribution for the WB and PCE in blue and orange, respectively. The thick, black, vertical line shows the KS D-statistic, the maximum distance between the two distributions. | 2012.12997 | cumulative_temp_mdf.jpg |
0.52076 | 7277a5c7780745e8a3cd634c49c1f8d0 | The calculated orbit for 2M14244053+4929580 aged 5 Gyr backwards from the present day, shown in the Galactic Cartesian coordinate system with a 1:1 aspect ratio in both projections to emphasize the planar nature of the orbit. The red dot marks the system's current location. It is clear that this system has a very disk-like orbit despite being having a metallicity ([Fe/H] ∼ -1.4) rather typical of the Galactic halo. | 2012.12997 | new_orbit_plot.jpg |
0.477829 | e72a0219ac6e43228ac7957fa181de23 | 2012.13002 | BE-basis.jpg |
|
0.473163 | 9cca6bb4438244b5b681b268fbc4fcba | 2012.13002 | BE-error-compare-1e-3-s.jpg |
|
0.509022 | e647fd8f624649f8996d30144c0b4901 | 2012.13002 | BE-error-compare-1e-3.jpg |
|
0.405002 | 5f1c115a11764449b62878dad4833184 | 2012.13002 | BE-s-basis.jpg |
|
0.459012 | 4ca15271bbe04a66ac8a62cf8da11421 | Average speedup (<ref>, excluding the cost of CP-ALS) of CCSD with rCP-DF-approximated PPL vs CP rank R (in units of the fitting basis, X) for the S66/12 dataset. The error bars denote the max/min speedup. | 2012.13002 | BE-time-nocp-1e3.jpg |
0.436097 | a492bea0c13d49749b00e9d0f075f5a4 | Average speedup (<ref>) of CCSD with rCP-DF-approximated PPL vs CP rank R (in units of the fitting basis, X) for the S66/12 dataset. ALS precision fixed at ϵ = 10^-3. The error bars denote the max/min speedup. | 2012.13002 | BE-time-w-1e3.jpg |
0.446617 | 814ffe195403473aa5c2115f0528900b | 2012.13002 | RE-s-error-1e3.jpg |
|
0.504469 | d7992431b7744d41b185065914d28bcb | 2012.13002 | RE-us-error-1e3.jpg |
|
0.403054 | fedf73e3899c442088bd7891ee687d45 | Average speedup (<ref>) of CCSD with rCP-DF-approximated PPL vs CP rank R (in units of the fitting basis, X) for the 7 largest clusters in the S66/12 dataset. The error bars denote the max/min speedup. | 2012.13002 | big-time-1e3.jpg |
0.385832 | 604562ee42224b5bb7addde213ba1101 | 2012.13002 | converge-test-DF.jpg |
|
0.442674 | 6958dc3a663c46248c4c2ad6a04b6e3f | 2012.13002 | converge-test-PS.jpg |
|
0.423713 | 18da134108c740caa28d49b0edfe6e31 | 2012.13002 | converge-test-rCP.jpg |
|
0.447763 | 1f53f64d9eea4bc8a5fa80163779cef9 | Absolute errors in matrix elements of g_ab,cd for a water dimer with S66 configuration approximated by the CP-PS, CP-DF, and rCP-DF factorizations obtained with ALS precision of ϵ = 10^-3. The error bars denote the max/min unsigned errors. | 2012.13002 | element-error.jpg |
0.430735 | c6d6d346b9424b13a9c28f6a32aa0119 | Percent of the total CCSD time spent in ALS for each cluster molecule in S66 dataset using rCP-DF with CP rank R = 1.3X and ALS precision of ϵ = 10^-3. Molecules are ordered according to the number of occupied orbitals, from smallest to largest. | 2012.13002 | percent_als.jpg |
0.446231 | bd6743429ba74c01977e63f2480db707 | Unsigned errors in the S66 CCSD binding energies (kcal/mol), relative to canonical CCSD, induced by the rCP-DF approximation to PPL. CP rank and ALS precision are fixed at R = 1.3X and ϵ = 10^-3, respectively. Molecules ordered from smallest to largest number of occupied orbitals. The orange line is the target maximum error, 0.1 kcal/mol, and the green line is the average error of the set. | 2012.13002 | s66-error.jpg |
0.464705 | 697dbc67c2c9402482e8faddad85e96e | Speedup (<ref>) of CCSD with rCP-DF-approximated PPL for the entire S66 dataset. CP rank and ALS precision are fixed at R = 1.3X and ϵ = 10^-3, respectively. Molecules are ordered according to the number of occupied orbitals, from smallest to largest. The orange line represents no speedup over CCSD and the green line is average speedup of the set. | 2012.13002 | s66-time.jpg |
0.501377 | 5fb1a71af5874f1cba5c4f0312eedcac | Average speedup (<ref>) of CCSD with rCP-DF-approximated PPL vs CP rank R (in units of the fitting basis, X) for the S66/7 dataset. ALS precision fixed at ϵ = 10^-3. The error bars denote the max/min speedup | 2012.13002 | time-basis-w.jpg |
0.470616 | f62f9aa6966f42e983a6362f8c339df8 | Ventilation performance. a) Measurement of tidal volume (red) and peak pressure (blue) over percentage of maximal displacement of the moving arm in ol-cmv mode. The red line indicates a linear dependence of the tidal volume on the arm movement. The attained peak pressure reflects the compliance of the text lung. b) Volume V (red), pressure p (blue) and flow Φ (orange) over time in pc-cmv mode for three different set pressures of 25 (dark blue), 20 (light blue) and 15 (bright blue) shown as dashed lines. c) Ventilation in vc-cmv mode for three different set flows of 30 (dark orange), 24 (light orange) and 18 (bright orange) shown as dashed lines. For b) and c) a PEEP of 7 was set and BPM is 10. | 2012.13005 | fig_data_211220.jpg |
0.411228 | 36d625dc2b364a819bb37ec193d6aa0b | Electronics design. a) Block diagram showing the power electronics, sensors, and user controls. The device is controlled by an Arduino Mega microcontroller. Connections for a 24 battery or for a second, redundant power supply as well as optional features, such as temperature monitoring and heating of the air/oxygen supply, are already implemented on the board. b) Photographs showing the front and back of the custom-made board, which is mounted directly to the front panel of the HDvent unit. | 2012.13005 | fig_electronics_211220.jpg |
0.459061 | c5cbf519e8594140a111cc1b1143b09b | Mechanical design of the ventilator. a) CAD model of the main components. A stepper motor (1) mounted to the primary support plate (2) drives a cam (3) and the movable arm (4) against the fixed arm (5). b) Photograph showing the mounting of the stepper motor and c) the mechanical arms in closed and open positions. | 2012.13005 | fig_mechanics_211220.jpg |
0.458909 | 93dfb7a418a44a32a589e6ac1126cfb8 | Overview of the HDvent system. a) Photograph showing the device housing a stepper motor as well as control electronics and sensors on a single PC board mounted to the front panel. Air is supplied to the patient via compression of the externally mountable ambu and standard hoses and valves. Ventilation parameters and alarm thresholds are accessible via the user interface on the front panel. b) Schematic of the HDvent architecture. All components directly in contact with the patient are single-use (orange). An optional monitoring unit (red) can be attached to the main device (blue) to display ventilation parameters, alarm thresholds and time traces. | 2012.13005 | fig_ventilator_211220.jpg |
0.405213 | 86048e7cbb40467aa367355f9ead7c8b | Monitoring unit hardware components and data flow. a) Data are transferred from the Arduino controller to the Raspberry Pi via a unidirectional serial link. b) On the Raspberry Pi, the data are processed, stored and a dashboard on an external display is generated, c). (1) time traces (30) of inspiratory pressure, flow and volume, (2) ventilator controller settings, (3) numerical indicators for the remaining diagnostics, (4) current ventilation mode with short description | 2012.13005 | ventilator_monitor_fig_v2.jpg |
0.48589 | 06985940b0d54cad9534dc8d0d40fa27 | A history of char/word error rates (CER/WER) on various ASR tasks in ESPnet. | 2012.13006 | asr_v2.jpg |
0.413558 | b2586b5120334c1894d55ebea131f430 | Potential Shapes f_1(r) = -1e^1.001r-1 full line and f(r) = -e^-arr dashed lines, where a = 0.5 was applied. | 2012.13008 | pot-yuk-hul-upper.jpg |
0.518954 | fbaccb8138944a61b9282e817f5c139d | Graphs for v_1(E), v(E), and v_2(E) corresponding to f_1(r) = -1e^1.001r-1, f(r) = -e^-0.5rr, and f_2(r) = -1e^0.966r-1 respectively, for -1 < E < 1. | 2012.13008 | spectral-hul-yuk-hul.jpg |
0.398462 | b15c582291f14bee9791a5620ffa4953 | (a) Crystal structure of YMn_6Sn_6, belonging to space group P6/mmm (191). Single-crystal neutron diffraction data showing the temperature dependence of the magnetic (b) wavevectors and (c) Bragg peak intensities in zero-field conditions. The inset of (c) shows the progression of the magnetic structures just below T_N ≈ 340 K. One commensurate structure emerges at T_N, but it is short lived and quickly gives way with decreasing temperature to two incommensurate structures. The legend for (c) is the same as in (b). (d) Magnetic structure phase diagram for the applied field, 𝐇, in-plane. The neutron diffraction experiments in this report were carried out with 𝐇∥[1̅, 1, 0]. The phases DS, TCS, FL, CAF, I, and FF correspond to distorted spiral, transverse conical spiral, fan-like, canted antiferromagnet, phase I, and forced ferromagnetic, respectively. | 2012.13010 | Fig1.jpg |
0.444734 | 654e0b1dff8f46368e1abbd5ee6fc144 | Energy comparison between “model 1” (orange) and “model 2” (blue) as spins in layers 3, 4, 7, and 8 deviate from the δ=0^∘ axis/the applied field direction. The other spins are either ± 68^∘ from the applied field direction, as determined by the Rietveld refined structures. Two values for | J_2/J_1 | are reported, where solid lines represent | J_2/J_1 | = 0.56 (full model), and dashed lines represent | J_2/J_1 | = 0.36 (reduced model). In both cases, the energy increases linearly as the spins deviate from the applied field direction for “model 2,” whereas the total energy initially decreases for “model 1.” | 2012.13010 | Fig10.jpg |
0.398396 | 4359422cd041480c9612e79fceff2aeb | The coordinate system used in calculating the magnetic structure factors, where x, y, and z are unit vectors which define a right-handed Cartesian coordinate system. The vector, S⃗_n, is the spin magnitude and direction for Mn moments in layer n, and γ and δ are the values defined by the angle that S⃗_n makes with the applied field direction, H⃗. | 2012.13010 | Fig11.jpg |
0.402966 | cc5a5ebb7eb9420cbf782e8eec2fd9d8 | The observed versus calculated magnetic structure factor squared results for the magnetic 'model 2' for the low temperature high-field phase. The calculated result was obtained via Rietveld refinement for the magnetic intensity only. Magnetic Bragg peaks have wavevectors, (0, 0, 0), (0, 0, 0.25), and (0, 0, 0.5). Data were taken at 1.5 K and 7.8 T in the (H, H, L) scattering plane. | 2012.13010 | Fig12.jpg |
0.411604 | 4683102f67ea4b8c94ac0969de80e3b3 | Room temperature data showing the evolution of the magnetic structures with increasing applied field. (a) The change in periodicity for the two incommensurate structures with wavevectors, (0, 0, k_z,n). (b) The change in intensity for magnetic Bragg peaks about the (0, 0, 2) reciprocal lattice point. As the Bragg peak at (0, 0, 2-k_z,2) rapidly decreases in intensity above 2 T, the commensurate Bragg peak at (0, 0, 2.5) just as rapidly increases in intensity. The dashed lines for both data sets are to emphasize the relationship between the two, which suggests the incommensurate k_z,2 structure is transitioning into the commensurate structure above 2 T. | 2012.13010 | Fig2.jpg |
0.458494 | 0d68f77465cd462bb3bcb26c0613c888 | Results of the best fit magnetic model for the room temperature, low-field canted antiferromagnetic phase with wavevectors, (0, 0, 0) and (0, 0, 0.5). Data were taken at 295 K and 2 T in the (H, H, L) scattering plane, and the figures represent the magnetic structure at this particular field and temperature. (a) The observed versus calculated magnetic structure factor squared. The calculated result was obtained via Rietveld refinement for the magnetic intensity only. The inset shows the magnetic structure in the ab-plane. Only Mn ions are shown (grey spheres), and the orange and blue arrows represent the two different directions of the moments. (b) The magnetic unit cell, where the dashed line defines the size of nuclear unit cell. | 2012.13010 | Fig3.jpg |
0.480735 | ddf9b1fb04aa4a7b8ef617b3cb8144f7 | Results for “model 1” of the low temperature, high-field fan-like phase with wavevectors, (0, 0, 0), (0, 0, 0.25), and (0, 0, 0.5). Data were taken at 1.5 K and 7.8 T in the (H, H, L) scattering plane. (a) The observed versus calculated magnetic structure factor squared. The calculated result was obtained via Rietveld refinement for the magnetic intensity only. The inset shows the refined magnetic structure in the ab-plane for 1.5 K and 7.8 T, where γ = 68^∘± 2^∘ and δ = 0^∘± 1^∘. Only Mn ions are shown (grey spheres), and the orange and blue arrows represent the two different angle magnitudes, γ and δ, respectively, which define the moment directions away from the applied field direction. (b) The magnetic unit cell, where the dashed lines define the size of the nuclear unit cell. | 2012.13010 | Fig4.jpg |
0.433118 | 413938f1a4ea4b4283b6903decbda776 | High-field data with 𝐇∥[1̅, 1, 0], taken at 10 K with a position sensitive detector and moderately course resolution. Between the FL phase (9.0 T and 9.5 T) and FF phase (10.5 T), an intermediate phase appears, where the 0.5-like peaks disappear, and the magnetic structure is incommensurate. | 2012.13010 | Fig5.jpg |
0.470775 | 479881f4e0254113adc9590f955b3716 | Results from the analyzed synchrotron powder diffraction data. (a) The lattice parameters and (b) unit cell volume versus temperature. (c) Rietveld refinement for 295 K data. The data are displayed as black dots and the Rietveld calculated fit is the solid yellow line running through the data. The top row of tik marks (green) denote YMn_6Sn_6 Bragg peak positions, and the bottom row of tik marks (red) are the elemental Sn impurity Bragg peak positions. The difference curve (observed-calculated) is shown as the solid blue line on the bottom of the plot. The inset highlights the fit in the high-Q region. | 2012.13010 | Fig6.jpg |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.