Datasets:

xvjiarui

/

arxiv_mini

like 1

Fingerprints: (a) Left (b) Right (c) Arch (d) Whorl. Fingerprints of class arch are not considered.

insidecity_art1677

street_a232022

street_urb104

Raw images, 2D data extracted from the images, and overlaid\n computed 3D position and body orientation of a hummingbird ({\\em\n Calypte anna}). In these images, a blue circle is drawn centered\n on the 2D image coordinates $(\\tilde{u},\\tilde{v})$. The blue line\n segment is drawn through the detected body axis ($\\theta$ in Section\n \\ref{data_association}) when eccentricity ($\\epsilon$) of the\n detected object exceeds a threshold. The orange circle is drawn\n centered on the 3D estimate of position $(x,y,z)$ reprojected\n through the camera calibration matrix $\\cammatrix$, and the orange\n line segment is drawn in the direction of the 3D body orientation\n vector.

Illustration of our schemes: Figure~\\ref{fig:OurScheme:origianltable} is the bank transaction screenshot which contains a sensitive transaction table to be protected. Figure~\\ref{fig:OurScheme:MS1} illustrates method 1 where the sensitive table is replaced by barcodes; and the mobile device captures, verifies and decodes part of the table. Figure~\\ref{fig:OurScheme:MS23} illustrates method 2 where the sensitive table is displayed with barcodes; and the table is rendered on both terminal and mobile to be compared by the user. The decoded tables are generated by our proof-of-concept implementation which are then ``cut-and-paste'' to produce the illustration. The green boxes show the captured region and the red dots are for image registration.

Results on a real image. (\\textbf{A}) Noise-free immunofluorescence image of actin fibres (Courtesy of C. Aemisegger, CMIA, University of Z\xef\xbf\xbdrich); (\\textbf{B}) Image corrupted with additive Gaussian noise, $\\mathrm{PSNR}=12.20 \\ \\mathrm{dB}$; (\\textbf{C}) Isotropic smoothing, $\\mathrm{PSNR}=15.38 \\ \\mathrm{dB}$; (\\textbf{D}) Diffusion filtering, $\\mathrm{PSNR}=15.50 \\ \\mathrm{dB}$; (\\textbf{E}) Our algorithm, $\\mathrm{PSNR}=15.80 \\ \\mathrm{dB}$.

An example of a saliency map using Koch-Ullman model

Snapshots of model outputs

(Left subplot) Original image of a cross section of the brain; (Middle subplot) Blurred image; (Right subplot) Blurred and noisy image.

Image reconstruction results corresponding to Fig.\\ref{Fig4}: (Upper row of subplots) The MSE-optimal RL and RLTV estimates; (Lower row of subplots) supervised (kSVD) and unsupervised (splines) SRL recovery.