IPOL
http://www.ipol.im/feed/
IPOL Preprints — Latest public preprints from IPOL.An SEAIR model with personalised risk prediction scores and application to the Covid-19 epidemicOliver Boulant,
Theodoros Evgeniou,
Mathilde Fekom,
Anton Ovchinnikov,
Raphaël Porcher,
Camille Pouchol,
Nicolas Vayatishttp://www.ipol.im/pub/pre/305/
http://www.ipol.im/pub/pre/305/
Thu, 04 Jun 2020 15:57:59 +02002020-06-04T15:04:20ZThe aim of the present work is to provide an SEAIR framework which takes a personalised risk
prediction score as an additional input. Each individual is categorised depending on his actual
status with respect to the disease - mild or severe symptoms -, and the level of risk predicted
-low or high. This idea leads to a 4-fold extension of the ODE model in classical SEAIR. This
model offers the possibility for policy-makers to explore differentiated containment strategies,
by varying sizes for the low risk segment and varying dates for ‘progressive release’ of the
population, while taking into account the discriminative capacity of the risk score through its
AUC. Differential contact rates for low-risk/high-risk compartments are also included in the
model. The demo allows to select contact rates and time-depending exit strategies. The hard-
coded parameters correspond to the data for the Covid-19 epidemic in France, and the risk refers
to the probability of being admitted in ICU upon infection. Some examples of simulations are
provided.A Daily Measure of the SARS-CoV-2 Daily Reproduction Number for all CountriesTahar Zamene Boulmezaoud,
Luis Álvarez,
Miguel Colom,
Jean-Michel Morelhttp://www.ipol.im/pub/pre/304/
http://www.ipol.im/pub/pre/304/
Mon, 01 Jun 2020 17:49:45 +02002020-06-02T07:22:41ZWe propose a transparent method calculating every three hours for any country a 'daily reproduction number (DRN)' from the daily count of newly detected cases. The same method
applies to compute a daily reproduction number from the daily death count, which yields therefore another independent prediction of the expansion of the pandemic. A good experimental fit of both DRN curves is observed in many countries. Both curves are similar, with a delay that depends on each country’s detection and administrative processing delays. Both DRNs can be consulted daily online in the demo tag associated with this paper. We refer the readers to the online demo to experiment by themselves. The calculation of this DRN is based on a recent paper by Tahar Boulmezaoud.PALMS Image Partitioning Lab - A Toolbox for Image Partitioning with the
Piecewise Affine-Linear Mumford-Shah ModelLukas Kiefer,
Martin Storath,
Andreas Weinmannhttp://www.ipol.im/pub/pre/295/
http://www.ipol.im/pub/pre/295/
Wed, 06 May 2020 10:58:13 +02002020-05-06T09:13:52ZWe present a toolbox for computing approximate solutions of the piecewise affine-linear Mumford-
Shah model - PALMS Image Partitioning Lab. The piecewise affine-linear Mumford-Shah model
is a variational approach to image partitioning. The underlying algorithm is based on a splitting
approach using ADMM. The emerging subproblems are solved exactly and efficiently. We detail
the solver for these subproblems which is based on dynamic programming and incorporate an
acceleration strategy. The subproblems can be solved in parallel which our implementation
makes use of to provide an efficient overall algorithm. We conduct extended studies on the
effects of the algorithmic parameters. Thereby, the implemented algorithm is further optimized
w.r.t. runtime and efficiency. Finally, we underpin the efficiency of the algorithm by a compar-
ison with the state-of-the-art which shows that the presented algorithm has lower computation
times and yields lower mean functional values.An empirical algorithm to forecast the evolution of the number of COVID-19 symptomatic patients after social distancing interventionsLuis Álvarez,
Miguel Colom,
Jean-Michel Morelhttp://www.ipol.im/pub/pre/301/
http://www.ipol.im/pub/pre/301/
Sat, 25 Apr 2020 20:33:21 +02002020-06-02T07:22:41ZIn this work we present an empirical algorithm to forecast the evolution of the number of COVID-19 symptomatic patients after social distancing interventions. The algorithm is based on a low dimensional model for the variation of the exponential growth rate that decreases after the implementation of the social distancing measures. In addition, the incubation time of COVID-19 is taken into account in the analysis. An experimental work is carried out adjusting the model parameters to different countries such as China, South Korea, Italy, France and Spain, obtaining consistent results but which may be subject to significant errors due to the lack of reliability in the data used.Finding the Skeleton of 2D Shape and Contours: Implementation of Hamilton-Jacobi SkeletonYuchen He,
Sung Ha Kang,
Luis Álvarezhttp://www.ipol.im/pub/pre/296/
http://www.ipol.im/pub/pre/296/
Tue, 03 Mar 2020 16:12:45 +01002020-03-03T15:12:45ZThis paper presents the details of the flux-ordered thinning algorithm, which we refer to
as the Hamilton-Jacobi Skeleton (HJS). It computes the skeleton of any binary 2D shape.
This is based on the observation that the skeleton points have low average outward flux of the
gradient of the distance transform. The algorithm starts by computing the distance function and
approximating the flux values for all pixels inside the shape. Then a procedure called homotopy
preserving thinning iteratively removes points with high flux while preserving the homotopy of
the shape. In this paper, we implement the distance transform using a fast sweeping algorithm.
We present numerical experiments to show the performance of HJS applied to various shapes.
We point out that HJS serves as a multi-scale shape representation, a homotopy classifier,
and a deficiency detector for binary 2D shapes. We also quantitatively evaluate the shape
reconstructed from the medial axis obtained by HJS.An Algorithm for 3D Curve SmoothingDaniel Santana-Cedrés,
Nelson Monzón,
Luis Álvarezhttp://www.ipol.im/pub/pre/292/
http://www.ipol.im/pub/pre/292/
Thu, 13 Feb 2020 00:45:21 +01002020-02-12T23:45:21ZIn this article we present an application of variational techniques to the smoothing of 3D curves. We study 2 types of application scenarios: in the first one the curve is just given by an ordered set of 3D points and in the second one the curve represents the medial axis of a 3D volume. In this last scenario, the input of the algorithm is the 3D volume and a 3D curve representing an approximation of the volume medial axis. We propose an algorithm for 3D curve smoothing, based on the minimization of a general variational model, which includes both scenarios. We present a variety of experiments to show the performance of the proposed technique.On Anisotropic Optical Flow Inpainting AlgorithmsLara Raad,
Maria Oliver,
Coloma Ballester,
Gloria Haro,
Enric Meinhardthttp://www.ipol.im/pub/pre/281/
http://www.ipol.im/pub/pre/281/
Mon, 03 Feb 2020 23:22:24 +01002020-02-03T22:28:47ZThis work describes two anisotropic optical flow inpainting algorithms. The first one recovers the missing flow values using the Absolutely Minimizing Lipschitz Extension partial differential equation (also called infinity Laplacian equation) and the second one uses the Laplace partial differential equation, both defined on a Riemmanian manifold. The Riemannian manifold is defined by endowing the plane domain with an appropriate metric depending on the reference video frame. A detailed analysis of both approaches is provided and their results are compared on three different applications: flow densification, occlusion inpainting and large hole inpainting.Image Inpainting using Patch Consensus and DCT PriorsIgnacio Ramírez Paulino,
Ignacio Houniehttp://www.ipol.im/pub/pre/286/
http://www.ipol.im/pub/pre/286/
Tue, 03 Dec 2019 16:48:20 +01002019-12-03T15:48:20ZWe present an implementation of the PACO-DCT inpainting algorithm. This method is based
on minimizing the likelihood of image patches in terms of their DCT coefficients, while requiring
consensus on the overlapping patches. The resulting problem is solved as an instance of the
PACO framework.An 'All Terrain' Crack Detector Obtained by Deep Learning on Available DatabasesSébastien Drouyerhttp://www.ipol.im/pub/pre/282/
http://www.ipol.im/pub/pre/282/
Mon, 28 Oct 2019 14:48:29 +01002019-10-28T19:01:01ZWe present a general deep learning method for detecting cracks on all sorts of surfaces. For
making this method robust to different types of cracks and acquisition procedures, we have
trained our method on four datasets - Crack500, DeepCrack, SDNet2018 and CrackForest. We
have also labelled the SDNet2018 dataset so that it contains semantic labels, as it originally
only proposed crack/non-crack classifications on the image level. To validate our approach, we
perform a cross-dataset study where we train the model on a subset of the datasets and test it
on another subset. Results of this study show that training the model on these various datasets
makes it more robust to new images, outperforming existing classical and deep learning methods.
In order to make our method even more robust to different objects, scenes and illuminations, we
have also added images from the Flickr website, leading to an important drop in false positives
on extra dataset images. The network seems to function well on images not belonging to any of
the datasets, and its publication in IPOL will allow users to enrich further training.Cloud Detection by Disparity Phase Analysis for Pushbroom Satellite ImagersTristan Dagobert,
Rafael Grompone von Gioi,
Carlo de Franchis,
Jean-Michel Morelhttp://www.ipol.im/pub/pre/271/
http://www.ipol.im/pub/pre/271/
Mon, 22 Jul 2019 00:27:49 +02002019-11-05T08:55:16ZThis paper proposes a cloud detector for earth observation images obtained by a push-broom satellite imager. This particular technology induces an inter-channel acquisition delay leading to a parallax effect for the clouds. We propose a method exploiting this characteristic thanks to the analysis of the visible inter-bands disparity maps. Several other features discriminating clouds are also defined and all are merged to build a robust a contrario statistical decision. Experiments applied on scenes acquired by various push-broom satellites such that Sentinel-2, Rapideye and Worldview-2 show the effectiveness of the proposed method. In particular, we demonstrate 97% and 98% success rates for cloud and non cloud classification for Sentinel-2 images.Cross-comparison of the Performance of Sequential Summed Area Table and Box Filter Algorithms with respect to C/C++ CompilersAli Ozturk,
Ibrahim Cayirogluhttp://www.ipol.im/pub/pre/268/
http://www.ipol.im/pub/pre/268/
Mon, 22 Jul 2019 00:03:34 +02002019-07-21T22:03:34ZSummed area table algorithm has been used to accelerate some computer vision and signal processing algorithms. In this study, the performance of the sequential summed area table algorithms and box filter algorithm with and without summed area table algorithm is examined by taking account of effect of C/C++ compilers. Three variants of sequential summed area table algorithm are included into the study. Loop invariant code motion and loop unrolling optimization techniques are applied to one of them. The performance of GNU Compiler Collection (GCC) and Intel C/C++ Compiler (ICC) on both Windows and Ubuntu and Visual C++ (CL) compiler on Windows by using the summed area table and box filter algorithms are compared. Result of the study reveals that Intel C/C++ Compiler (ICC) perform best on Ubuntu with respect to others for sequential summed area table and box filter algorithm. The performance of summed area table calculation which utilizes Viola-Jones Equation by using a scalar accumulator outperforms other summed area table algorithms.Visibility Detector for Time Series of Spectrally Limited Satellite ImagersTristan Dagobert,
Jean-Michel Morel,
Carlo de Franchishttp://www.ipol.im/pub/pre/245/
http://www.ipol.im/pub/pre/245/
Fri, 19 Jul 2019 18:25:37 +02002019-07-20T15:20:39ZThis article addresses the problem of estimating scene visibility in time
series of satellite images. We are especially focused on satellites with few
spectral bands but with high revisit frequency. Our approach exploits the
redundancy of information acquired during these revisits. It is based on
an unsupervised algorithm that tracks local ground textures across time
and detects ruptures caused by opaque clouds, haze, cirrus and shadows.
Experiments have been carried out on 18 Planet times series representing
various locations. These time series come with hand-made labeled ground
truth that we make available to the scientific community. We compare the
results obtained with those of the PlanetScope algorithm and demonstrate
the effectiveness of the proposed method: success rates of 94% and 84% are
reached for the visible and occulted regions classification.A Parallel, O (n) Algorithm for an Unbiased, Thin WatershedThéodore Chabardès,
Petr Dokládal,
Matthieu Faessel,
Michel Bilodeauhttp://www.ipol.im/pub/pre/215/
http://www.ipol.im/pub/pre/215/
Thu, 09 May 2019 12:30:32 +02002019-05-09T10:30:32ZThe watershed transform is a powerful tool for morphological segmentation. Most common
implementations of this method involve a strict hierarchy on gray tones in processing the pixels
composing an image. This hierarchical dependency complexifies the efficient use of modern
computational architectures. This paper introduces a new way of computing the watershed
transform that alleviates the sequential nature of hierarchical queue propagation. It is shown
that this method can directly relate to the hierarchical flooding. Simultaneous and disorderly
growth can now be used to maximize performances on modern architectures. Higher speed is
reached, bigger data volume can be processed. Experimental results show increased performances
regarding execution speed and memory consumption.