Search

The detection of intracranial aneurysms from Magnetic Resonance Angiography images is a problem of rapidly growing clinical importance. In the last 3 years, the raise of deep convolutional neural networks has instigated a streak of methods that have shown promising performance. The major issue to address is the very severe class imbalance. Previous authors have focused their efforts on the network architecture and loss function. This paper tackles the data. A rough but fast annotation is considered: each aneurysm is approximated by a sphere defined by two points. Second, a small patch approach is taken so as to increase the number of samples. Third, samples are generated by a combination of data selection (negative patches are centered half on blood vessels and half on parenchyma) and data synthesis (patches containing an aneurysm are duplicated and deformed by a 3D spline transform). This strategy is applied to train a 3D U-net model, with a binary cross entropy loss, on a data set of 111 patients (155 aneurysms, mean size 3.86mm ± 2.39mm, min 1.23mm, max 19.63mm). A 5-fold cross-validation evaluation provides state of the art results (sensitivity 0.72, false positive count 0.14, as per ADAM challenge criteria). The study also reports a comparison with the focal loss, and Cohen's Kappa coefficient is shown to be a better metric than Dice for this highly unbalanced detection problem., La détection d'anévrismes intracrâniens à partir d'images d'angiographie par résonance magnétique est un problème d'une importance clinique croissante. Au cours des 3 dernières années, l'augmentation des réseaux de neurones convolutifs profonds a suscité une série de méthodes qui ont montré des performances prometteuses. Le problème majeur à résoudre est le très grave déséquilibre de classe. Les auteurs précédents ont concentré leurs efforts sur l'architecture du réseau et la fonction de perte. Cet article aborde les données. Une annotation grossière mais rapide est envisagée : chaque anévrisme est approché par une sphère définie par deux points. Deuxièmement, une approche par petits patches est adoptée afin d'augmenter le nombre d'échantillons. Troisièmement, les échantillons sont générés par une combinaison de sélection de données (les patches négatifs sont centrés à moitié sur les vaisseaux sanguins et à moitié sur le parenchyme) et la synthèse de données (les patches contenant un anévrisme sont dupliqués et déformés par une transformation spline 3D). Cette stratégie est appliquée pour former un modèle U-net 3D, avec une perte d'entropie croisée binaire, sur un ensemble de données de 111 patients (155 anévrismes, taille moyenne 3,86 mm ± 2,39 mm, min 1,23 mm, max 19,63 mm). Une évaluation de validation croisée 5 fois fournit des résultats de pointe (sensibilité 0,72, nombre de faux positifs 0,14, selon les critères de défi ADAM). L'étude rapporte également une comparaison avec la perte focale, et le coefficient Kappa de Cohen s'avère être une meilleure métrique que Dice pour ce problème de détection très déséquilibré.

International audience; In this paper, we propose a novel accelerated alternating optimization scheme to solve block-biconvex nonsmooth problems whose objectives can be split into smooth (separable) regularizers and simple coupling terms. The proposed method performs a Bregman distance based generalization of the well-known forward-backward splitting for each block, along with an inertial strategy which aims at getting empirical acceleration. We discuss the theoretical convergence of the proposed scheme and provide numerical experiments on image colorization.

International audience; Active contours (also known as snakes) have shown their ability to introduce regularity on image segmentation. In contrast with level-set approaches, the active contours techniques based on a contour parameterization are able to maintain the initial topology of the area of interest. For this reason, it has been used in recent medical research for diaphragm segmentation. Most of the on-line codes for 2D/3D segmentation, as well as built-in Matlab toolboxes are based on level-set methods. Moreover, in the literature, the implementation details of active contours methods with meshes in three dimensions are tight, making tedious any reproduction of these techniques. In this paper, we give some details of the implementation of active contours in 2D/3D with meshes, especially about the choice of the use of a 2D/3D mesh and its refinement. We also explore the choice of the parameters with a quantitative study of their influence on the segmentation results. The 3D segmentation method has been applied to CT scan images of the lungs.

International audience; We propose to detect defects in old movies, as the first step of a larger framework of old movies restoration by inpainting techniques. The specificity of our work is to learn a film restorer's expertise from a pair of sequences, composed of a movie with defects, and the same movie which was semi-automatically restored with the help of a specialized software. In order to detect those defects with minimal human interaction and further reduce the time spent for a restoration, we feed a U-Net with consecutive defective frames as input to detect the unexpected variations of pixel intensity over space and time. Since the output of the network is a mask of defect location, we first have to create the dataset of mask frames on the basis of restored frames from the software used by the film restorer, instead of classical synthetic ground truth, which is not available. These masks are estimated by computing the absolute difference between restored frames and defectuous frames, combined with thresholding and morphological closing. Our network succeeds in automatically detecting real defects with more precision than the manual selection with an all-encompassing shape, including some the expert restorer could have missed for lack of time.

International audience; This paper aims to couple the powerful prediction of the convolutional neural network (CNN) to the accuracy at pixel scale of the variational methods. In this work, the limitations of the CNN-based image colorization approaches are described. We then focus on a CNN which is able to compute a statistical distribution of the colors for each pixel of the image based on a learning over a large color image database. After describing its limitation, the variational method of [17] is briefly recalled. This method is able to select a color candidate among a given set while performing a regularization of the result. By combining this approach with a CNN, we designed a fully automatic image coloriza-tion framework with an improved accuracy in comparison with CNN alone. Some numerical experiments demonstrate the increased accuracy performed by our method.

International audience; We propose a new video inpainting model for movies restoration application. Our model combines structural reconstruction with a diffusion-based method and textural reconstruction with a patch-based method. Both proposed energies (one for each method) are alternatively minimized in order to preserve the overall structure while adding textural refinement. While the structural reconstruction is obtained jointly with optical flow computation with several proximal approaches, the textural reconstruction is processed by a variational non-local approach. Preliminary results on different Middlebury frames show quality improvement in the reconstruction.

International audience; This paper focuses on the denoising of chrominance channels of color images. We propose a variational framework involving TV reg-ularization that modifies the chrominance channel while preserving the input luminance of the image. The main issue of such a problem is to ensure that the denoised chrominance together with the original luminance belong to the RGB space after color format conversion. Standard methods of the literature simply truncate the converted RGB values, which lead to a change of hue in the denoised image. In order to tackle this issue, a " RGB compatible " chrominance range is defined on each pixel with respect to the input luminance. An algorithm to compute the orthogonal projection onto such a set is then introduced. Next, we propose to extend the CLEAR debiasing technique to avoid the loss of colourfulness produced by TV regularization. The benefits of our approach with respect to state-of-the-art methods are illustrated on several experiments.

Colorization of gray-scale images relies on prior color information. Exemplar-based methods use a color image as source of such information. Then the colors of the source image are transferred to the gray-scale target image. In the literature, this transfer is mainly guided by texture descriptors. Face images usually contain few texture so that the common approaches frequently fail. In this paper, we propose a new method taking the geometric structure of the images rather their texture into account such that it is more reliable for faces. Our approach is based on image morphing and relies on the YUV color space. First, a correspondence mapping between the luminance Y channel of the color source image and the gray-scale target image is computed. This mapping is based on the time discrete metamorphosis model suggested by Berkels, Effland and Rumpf. We provide a new finite difference approach for the numerical computation of the mapping. Then, the chrominance U,V channels of the source image are transferred via this correspondence map to the target image. A possible postprocessing step by a variational model is developed to further improve the results. To keep the contrast special attention is paid to make the postprocessing unbiased. Our numerical experiments show that our morphing based approach clearly outperforms state-of-the-art methods.

International audience; This paper deals with the difficult problem of video colorization. Methods in the literature are generally based on spatio-temporal video blocks, or on frame to frame color propagation methods, each technique having its own advantages and drawbacks. In this paper , we present both a novel automatic frame-to-frame propagation approach and an interactive correction method within a variational framework. The proposed method propagates colors from an initial colorized frame to the whole grayscale video sequence. The automatic propagation results may be visually unsuitable in some cases. To overcome this limitation, a spatio-temporal functional with a user-guided correction is introduced. Two primal-dual algorithms are designed to solve the proposed variational models. Numerical results show the efficiency and the potentiality of the proposed approach in comparison with state-of-the-art methods.

International audience; The aim of this paper is twofold. First, we propose a new method for enhancing the contrast of gray-value images. We use the difference of the average local contrast measures between the original and the enhanced images within a variational framework. This enables the user to control the contrast level and the scale of the enhanced details intuitively. Moreover, our model avoids large modifications of the original image histogram. Thereby it preserves the global illumination of the scene and can cope with large areas having similar gray values. The minimizer of the proposed functional is computed by a gradient descent algorithm in connection with a polynomial approximation of the average local contrast measure. The polynomial approximation is done via Bernstein polynomials and leads to to a speed up of the algorithm by applying fast Fourier transforms. In the second part, the approach is extended to a variational enhancement method for color images. The model approximately preserves the hue of the original image and includes additionally a total variation term to correct possible noise. The method requires no post-or pre-processing. The minimization problem is solved with a hybrid primal-dual algorithm. Numerical experiments demonstrate the efficiency and the flexibility of the proposed approaches in comparison with state-of-the-art methods.

International audience; This paper proposes a novel model for contrast enhancement of RGB images. The average local contrast measure is increased within a variational framework which preserves the hue of the original image by coupling the channels. The user is enabled to intuitively control the level of the contrast as well as the scale of the enhanced details. Moreover, our model avoids large modifications of the original image his-togram and thereby preserves the global illumination of the scene. The minimizer of the proposed functional is computed by a hybrid primal-dual algorithm. Numerical experiments show the reliability of the proposed approach in comparison with state-of-the-art methods.

With the advent of the microelectronics and the "smart-devices", the need for micro-sources of energy has risen. To meet these needs, a massive production of microbatteries is required in the next decades. Physical vapor deposition (PVD) is the process commonly used to fabricate thin film all solid state microbatteries. Commercially, LiCoO2 and lithium are used respectively as positive and negative electrodes. LiPON is commonly used as solid electrolyte because of its good ionic conductivity, good thermal and electrochemical stability and also for its good interfacial stability with the electrodes. However, the PVD process requires very expensive deposition machineries, with a high energy cost, which impact consequently the price and the development of the microbatteries. Processes involving liquid step, such as ink jet printing or screen printing propose cheaper machinery and much higher production rate. Moreover, these techniques are precise enough to deposit any pattern either by located drop jetting or by use of specific mask fitting the desired shape. In this communication, we will report our efforts to develop all or partially printed microbatteries. We have worked on replacing LiPON by another solid electrolyte which could be deposited by printing. Ionogels are solid electrolyte in which an ionic liquid (IL) is confined within a solid host network. Although solid like electrolytes, ionogels are formed from a liquid precursor, and are thus appropriate materials for the purpose. The liquid precursor is obtained by mixing the IL with organic, inorganic or hybrid monomers that are then polymerized to build in-situ the confining network. Furthermore, ionogels show good ionic conductivity and also good thermal and electrochemical stability, thus making them attractive solid electrolyte for lithium batteries. The time required to form a fully solidified material can vary depending on the nature of the confining network and the polymerization process. By using photo-polymerization, which is very fast, the manufacturing time is lesser than an hour, while for LiPON it can last more than ten hours. We have identified a promising printable photo-polymerizable ionogel composition with high ionic conductivity, pressure-sensitive-adhesive character, low interface resistance with lithium electrode, which allows stable cycling of LiFePO4/lithium cells, at least up to 1200 cycles at ambient temperature, C/5 rate and for surface capacities typical of microbatteries, e.g. 0.15 mAh cm-2. This ionogel is based on short molecular weight monomers with high content of polymerizable functionalities (triacrylate and diacrylate) and ILs composed of fluorinated sulfonylimide anions combined with alkylated pyrrolidinium cations [1]. We have evaluated the feasibility of making all or partially printed microbatteries with this printable solid electrolyte, not starting from the ground but trying to fit in an existing industrial process that uses PVD for manufacturing microbatteries. Microbatteries assembled from a dense PVD made LiCoO2 layer, a printed photo-polymerized ionogel, and a lithium layer exhibit good performance. However, the photo-polymerized ionogel showed very low compatibility when the lithium electrode is directly deposited by PVD on the top of it. Thus the lithium electrode must be mechanically assembled with the ionogel electrolyte. Microbatteries assembled from a printed porous LiFePO4 composite layer [2], a printed photo-polymerized ionogel, and lithium layer (an already made foil or evaporated layer and mechanically pressed on the ionogel surface) also demonstrated good performance. Acknowledgements Financial funding from the project TOURS 2015 n° O12590-418393 (Programme d’Investissement d’Avenir – FSN – AAP Nanoélectronique n°1) is acknowledged. References [1] D. Aidoud, A. Etiemble, D. Guy-Bouyssou, E. Maire, J. Le Bideau, D. Guyomard, B. Lestriez, "Interfacial stability and electrochemical behavior of Li/LiFePO4 microbatteries using novel soft and weakly adhesive photo-ionogel electrolytes”, J. Power Sources, 330, 2016, 92-103. [2] P.-E. Delannoy, B. Riou, T. Brousse, J. Le Bideau, D. Guyomard, B. Lestriez, “Ink-jet printed porous composite LiFePO4 electrode from aqueous suspension for microbatteries”, J. Power Sources, 2015, 287, 261-268.

International audience

International audience; This paper is concerned with a problem arising when editing color images, namely the Luminance-Hue Specification. This problem often occurs when converting an edited image in a given color-space to RGB. Indeed, the colors often get out of the standard range of the RGB space which is commonly used by most of display hardwares. Simple truncations lead to inconsistency in the hue and luminance of the edited image. We formalize and describe this problem from a geometrical point of view. A fast algorithm to solve the considered problem is given. We next focus on its application to image colorization in the RGB color-space while most of the methods use other ones. Using directly the three RGB channels, our model avoids artifact effects which appear with other color-spaces. Finally a variational model that regularizes color images while dealing with Luminance-Hue Specification problem is proposed.

International audience; La Patagonie depuis le XVe siècle attire les Européens. Cette attirance résulte de divers motifs avec une évolution dans le temps et dans l'espace. La Patagonie est un lieu propice aux utopies. Celle de la « Cité des Césars » (« Ciudad de los Cesares ») de la période coloniale, de « la Nouvelle Galle » du pasteur Jones, du royaume d'Araucanie d'Antoine Aurélie Premier, mais aussi celles des chercheurs d'or, des sociétés agricoles et industrielles, et celle des colons dans le cadre de ladite « Loi de Colonisation » de 1937. Alors qu'Aysén semble entrer lentement dans une ère de « modernité » apparaissent dès les années 1980 de nouvelles « utopies » individuelles, celles de « philanthropes verts », étrangers et chiliens, d'entrepreneurs verts « éco-touristiques », en même temps que le concept de « Aysén Reserva de Vida ». Quelles sont les incidences de ces nouveaux imaginaires dans un contexte de développement touristique ? S'agit-il des révolutions dans le mode de penser et de développer la Patagonie chilienne ? « Aysén Réserve de Vie » est-il simplement un mouvement cherchant à contrecarrer l'ultralibéralisme chilien et la globalisation industrielle à l'oeuvre en Patagonie ? S'agit-il d'une nouvelle utopie ?

This paper addresses the topic of image colorization that consists in converting a gray-scale image into a color one. In the literature, there exist two main types of approaches to tackle this problem. The first one is the manual methods where the color information is given by some scribbles drawn by the user on the image. The interest of these approaches comes from the interactions with the user that can put any color he wants. Nevertheless, when the scene is complex many scribbles must be drawn and the interactive process becomes tedious and time-consuming. The second category of approaches is the exemplar-based methods that require a color image as input. Once the example image is given, the colorization is generally fully automatic. A limitation of these methods is that the example image needs to contain all the desired colors in the final result. In this paper, we propose a new framework that unifies these two categories of approaches into a joint variational model. Our approach is able to take into account information coming from any colorization method among these two categories. Experiments and comparisons demonstrate that the proposed approach provides competitive colorization results compared to state-of-the-art methods.

International audience; This paper deals with the problem of image colorization. A model including total variation regularization is proposed. Our approach colorizes directly the three RGB channels, while most existing methods were only focusing on the two chrominance channels. By using the three channels, our approach is able to better preserve color consistency. Our model is non convex, but we propose an efficient primal-dual like algorithm to compute a local minimizer. Numerical examples illustrate the good behavior of our algorithm with respect to state-of-the-art methods.

International audience; This paper provides a new method to colorize gray-scale images. While the computation of the luminance channel is directly performed by a linear transformation, the colorization process is an ill-posed problem that requires some priors. In the literature two classes of approach exist. The first class includes manual methods that need the user to manually add colors on the image to colorize. The second class includes exemplar-based approaches where a color image, with a similar semantic content, is provided as input to the method. These two types of priors have their own advantages and drawbacks. In this paper, a new variational framework for exemplar-based colorization is proposed. A nonlocal approach is used to find relevant color in the source image in order to suggest colors on the gray-scale image. The spatial coherency of the result as well as the final color selection is provided by a nonconvex variational framework based on a total variation. An efficient primal-dual algorithm is provided, and a proof of its convergence is proposed. In this work, we also extend the proposed exemplar-based approach to combine both exemplar-based and manual methods. It provides a single framework that unifies advantages of both approaches. Finally, experiments and comparisons with state-of-the-art methods illustrate the efficiency of our proposal. 1. Introduction. The colorization of a gray-scale image consists of adding color information to it. It is useful in the entertainment industry to make old productions more attractive. The reverse operation is based on perceptual assumptions and is today an active research area [28], [13], [37]. Colorization can also be used to add information in order to help further analysis of the image by a user (e.g., sensor fusion [43]). It can also be used for art restoration ; see, e.g., [17] or [41]. It is an old subject that began with the ability of screens and devices to display colors. A seminal approach consists in mapping each level of gray into a color-space [18]. Nevertheless, all colors cannot be recovered without an additional prior. In the existing approaches, priors can be added in two ways: with a direct addition of color on

International audience; Color image enhancement is a challenging task in digital imaging with many applications. This paper contributes to image enhancement methods. We propose a new variational model for color improvement in the RGB space based on a desired target intensity image. Our model improves the visual quality of the color image while it preserves the range and takes the hue of the original, badly exposed image into account without amplifying its color artifacts. To approximate the hue of the original image we use the fact that affine transforms are hue preserving. To cope with the noise in the color channels we design a particular coupled TV regularization term. Since the target intensity of the image is unaltered our model respects important image structures. Numerical results demonstrate the very good performance of our method.

Towards the end of the 1980s a new idealization of nature, the wilderness, changed Patagonia into a "must" to discover and to save from the pressure of industrial exploitation. Through a constructivist approach, that we choose to define as bio-geographic, this article analyses the effects of the eco-philanthropic utopia of Douglas Tompkins in Patagonia. He builds a new aesthetic, utopian and personal, that provokes debates, beyond the territorial and political challenges that the sole issue of conservation and creation of private parks carries. Books, films and graphic works in favor of ecology and against industrialization and technology, changes views actors have on this remote land. Eco activists and ecotourism actors praise his contribution to tackle socio-environmental issues. The rural world, entrepreneurs and defenders of the spirit of the pioneers, clearers of virgin lands, criticize his opposition to development and his lack of respect for their way of life. Nevertheless, they sell their land to rich westerners and Chileans who want to possess their own private parks. Later, these actors, now landless, hope that tourism, strengthened by the creation of parks and the promotion of a pure nature, will ensure their economic future. With the rapid increase of private parks, the promotion of the "last wilderness and the creation of the biggest network of national parks of South America, the development paradigm of Patagonia is changing. The austral lands have become a valuable green treasure for tourists and landowners. It is yet to be understood if his will effectively benefit biodiversity protection and the wellbeing of its inhabitants., Al final de los años 80, una nueva idealización de la naturaleza, el "wilderness", transforma a la Patagonia en un lugar para disfrutar turísticamente y preservar de las codicias industriales. Desde una perspectiva constructivista que se propone denominar bio-geográfica, esta comunicación analiza la incidencia socio-espacial de la utopía eco-filantrópica de Douglas Tompkins in Patagonia. Al incentivar el resguardo de las bellezas naturales de la Patagonia y salvarlas de las codicias industriales, el instala un nuevo imaginario cultural y turístico. Moldea una nueva estética, utópica y personal, que genera debates, más allá de los desafíos territoriales y políticos que conlleva la conservación y creación de parques privados. Escritos, películas y afiches a favor de la ecología y en contra de la industrialización y la tecnologización de la sociedad, cambian la mirada de los actores locales y visitantes de estén rincón del mundo. Los activistas verdes y los emprendedores eco-turísticos alaban su contribución en las luchas socio-ambientales y protección de la biodiversidad. El mundo rural, empresarios y defensores del espíritu de los colonos, critican su oposición a su cultura y forma de desarrollo. Sin embargo, ellos venden sus tierras a los ricos norteamericanos, europeos y chilenos, deseosos de poseer sus propios parques. Luego, estos actores ya despojados de sus tierras, esperan que el turismo, incrementado por la creación de parques y promoción de una naturaleza prístina, les asegure su futuro económico. Con la rapida multiplicación de parques privados, la promoción del "ultimo wilderness" y la creación de la red de parques nacionales mas grande de sud-america, esta cambiando el paradigma del desarrollo de la Patagonia. Las tierras australe se han transformadas en un preciado tesoro verde tanto para turistas y proprietarios. Aún esta por entenderse si esto beneficiará realmente a la proteccion de la biodiversiad y el bienestar de sus habitantes