The development of an automatic telemedicine system for computer-aided screening and grading of diabetic retinopathy depends on reliable detection of retinal lesions in fundus images. In this paper, a novel method for automatic detection of both microaneurysms and hemorrhages in color fundus images is described and validated. The main contribution is a new set of shape features, called Dynamic Shape Features, that do not require precise segmentation of the regions to be classified. These features represent the evolution of the shape during image flooding and allow to discriminate between lesions and vessel segments. The method is validated per-lesion and per-image using six databases, four of which are publicly available. It proves to be robust with respect to variability in image resolution, quality and acquisition system. On the Retinopathy Online Challenge's database, the method achieves a FROC score of 0.420 which ranks it fourth. On the Messidor database, when detecting images with diabetic retinopathy, the proposed method achieves an area under the ROC curve of 0.899, comparable to the score of human experts, and it outperforms state-of-the-art approaches.

Figure 1: Given a reference arrangement composed of vector elements (top left), our analysis scheme divides the raw element set into appearance categories (bottom left). Spatial interactions based on appearance can be learned by statistical modeling and exploited to yield visually similar arrangements (right). AbstractWe present a technique for the analysis and re-synthesis of 2D arrangements of stroke-based vector elements. The capture of an artist's style by the sole posterior analysis of his/her achieved drawing poses a formidable challenge. Such by-example techniques could become one of the most intuitive tools for users to alleviate creation process efforts. Here, we propose to tackle this issue from a statistical point of view and take specific care of accounting for information usually overlooked in previous research, namely the elements' very appearance. Composed of curve-like strokes, we describe elements by a concise set of perceptually relevant features. After detecting appearance dominant traits, we can generate new arrangements that respect the captured appearance-related spatial statistics using multitype point processes. Our method faithfully reproduces visually similar arrangements and relies on neither heuristics nor post-processes to ensure statistical correctness.

Figure 1: Our Model-Based Discrete Texture Synthesis. Given a 2D or 3D exemplar texture composed of discrete vector elements, our model captures the pairwise element interactions that govern the texture's spatial organization, and accounts for their complex shapes. New output textures can then be generated to fill specified domain. AbstractWe present a novel shape-aware method for synthesizing 2D and 3D discrete element textures consisting of collections of distinct vector graphics objects. Extending the long-proven point process framework, we propose a shape process, a novel stochastic model based on spatial measurements that fully take into account the geometry of the elements. We demonstrate that our approach is well-suited for discrete texture synthesis by example. Our model enables for both robust statistical parameter estimation and reliable output generation by Monte Carlo sampling. Our numerous experiments show that contrary to current state-of-the-art techniques, our algorithm manages to capture anisotropic element distributions and systematically prevents undesirable collisions between objects.

International audienceThis paper targets two related color manipulation problems: Color transfer for modifying an image's colors and colorization for adding colors to a grayscale image. Automatic methods for these two applications propose to modify the input image using a reference that contains the desired colors. Previous approaches usually do not target both applications and suffer from two main limitations: possible misleading associations between input and reference regions and poor spatial coherence around image structures. In this paper, we propose a unified framework that uses the textural content of the images to guide the color transfer and colorization. Our method introduces an edge-aware texture descriptor based on region covariance, allowing for local color transformations. We show that our approach is able to produce results comparable or better than state-of-the-art methods in both applications