Publications

Extraction of ancient map contents using trees of connected components

Abstract

Ancient maps are an historical and cultural heritage widely recognized as a very important source of information, but exploiting such maps is complicated. In this project, we consider the Linguistic Atlas of France (ALF), built between 1902 and 1910. This cartographical heritage produces firstrate data for dialectological researches. In this paper, we focus on the separation of the content in layers for facilitating the extraction, the analysis, the visualization and the diffusion of the data contained in these ancient linguistic atlases.

Continue reading

A tutorial on well-composedness

Abstract

Due to digitization, usual discrete signals generally present topological paradoxes, such as the connectivity paradoxes of Rosenfeld. To get rid of those paradoxes, and to restore some topological properties to the objects contained in the image, like manifoldness, Latecki proposed a new class of images, called well-composed images, with no topological issues. Furthermore, well-composed images have some other interesting properties: for example, the Euler number is locally computable, boundaries of objects separate background from foreground, the tree of shapes is well-defined, and so on. Last, but not the least, some recent works in mathematical morphology have shown that very nice practical results can be obtained thanks to well-composed images. Believing in its prime importance in digital topology, we then propose this state-of-the-art of well-composedness, summarizing its different flavours, the different methods existing to produce well-composed signals, and the various topics that are related to well-composedness.

Continue reading

SmartDoc 2017 video capture: Mobile document acquisition in video mode

Abstract

As mobile document acquisition using smartphones is getting more and more common, along with the continuous improvement of mobile devices (both in terms of computing power and image quality), we can wonder to which extent mobile phones can replace desktop scanners. Modern applications can cope with perspective distortion and normalize the contrast of a document page captured with a smartphone, and in some cases like bottle labels or posters, smartphones even have the advantage of allowing the acquisition of non-flat or large documents. However, several cases remain hard to handle, such as reflective documents (identity cards, badges, glossy magazine cover, etc.) or large documents for which some regions require an important amount of detail. This paper introduces the SmartDoc 2017 benchmark (named “SmartDoc Video Capture”), which aims at assessing whether capturing documents using the video mode of a smartphone could solve those issues. The task under evaluation is both a stitching and a reconstruction problem, as the user can move the device over different parts of the document to capture details or try to erase highlights. The material released consists of a dataset, an evaluation method and the associated tool, a sample method, and the tools required to extend the dataset. All the components are released publicly under very permissive licenses, and we particularly cared about maximizing the ease of understanding, usage and improvement.

Continue reading

Derived-term automata of weighted rational expressions with quotient operators

Abstract

Quotient operators have been rarely studied in the context of weighted rational expressions and automaton generation—in spite of the key role played by the quotient of words in formal language theory. To handle both left- and right-quotients we generalize an expansion-based construction of the derived-term (or Antimirov, or equation) automaton and rely on support for a transposition (or reversal) operator. The resulting automata may have spontaneous transitions, which requires different techniques from the usual derived-term constructions.

Continue reading

Benchmarking keypoint filtering approaches for document image matching

Abstract

Reducing the amount of keypoints used to index an image is particularly interesting to control processing time and memory usage in real-time document image matching applications, like augmented documents or smartphone applications. This paper benchmarks two keypoint selection methods on a task consisting of reducing keypoint sets extracted from document images, while preserving detection and segmentation accuracy. We first study the different forms of keypoint filtering, and we introduce the use of the CORE selection method on keypoints extracted from document images. Then, we extend a previously published benchmark by including evaluations of the new method, by adding the SURF-BRISK detection/description scheme, and by reporting processing speeds. Evaluations are conducted on the publicly available dataset of ICDAR2015 SmartDOC challenge 1. Finally, we prove that reducing the original keypoint set is always feasible and can be beneficial not only to processing speed but also to accuracy.

Continue reading

The 4th reactive synthesis competition (SYNTCOMP 2017): Benchmarks, participants & results

Abstract

Continue reading

PaInleSS: A framework for parallel SAT solving

Abstract

Over the last decade, parallel SAT solving has been widely studied from both theoretical and practical aspects. There are now numerous solvers that dier by parallelization strategies, programming languages, concurrent programming, involved libraries, etc. Hence, comparing the eciency of the theoretical approaches is a challenging task. Moreover, the introduction of a new approach needs either a deep understanding of the existing solvers, or to start from scratch the implementation of a new tool. We present PaInleSS: a framework to build parallel SAT solvers for many-core environments. Thanks to its genericity and modularity, it provides the implementation of basics for parallel SAT solving like clause exchanges, Portfolio and Divide-and-Conquer strategies. It also enables users to easily create their own parallel solvers based on new strategies. Our experiments show that our framework compares well with some of the best state-of-the-art solvers.

Continue reading

Augmented songbook: An augmented reality educational application for raising music awareness

Abstract

This paper presents the development of an Augmented Reality mobile application which aims at sensibilizing young children to abstract concepts of music. Such concepts are, for instance, the musical notation or the concept of rythm. Recent studies in Augmented Reality for education suggest that such technologies have multiple benefits for students, including younger ones. As mobile document image acquisition and processing gains maturity on mobile platforms, we explore how it is possible to build a markerless and real-time application to augment the physical documents with didactical animations and interactive content. Given a standard image processing pipeline, we compare the performance of different local descriptors at two key stages of the process. Results suggest alternatives to the SIFT local descriptors, regarding result quality and computationnal efficiency, both for document model identification and pespective transform estimation. All experiments are performed on an original and public dataset we introduce here.

Continue reading

Caractérisation des zones de mouvement périodiques pour applications bio-médicales

Abstract

De nombreuses applications biomedicales impliquent l’analyse de séquences pour la caractérisation du mouvement. Dans cet article, nous considerons des séquences 2D+t où un mouvement particulier (par exemple un flux sanguin) est associé à une zone spécifique de l’image 2D (par exemple une artère). Mais de nombreux mouvements peuvent co-exister dans les séquences (par exemple, il peut y avoir plusieurs vaisseaux sanguins presents, chacun avec leur flux spécifique). La caractérisation de ce type de mouvement implique d’abord de trouver les zones où le mouvement est présent, puis d’analyser ces mouvements : vitesse, régularité, fréquence, etc. Dans cet article, nous proposons une méthode appropriée pour détecter et caractériser simultanément les zones où le mouvement est présent dans une séquence. Nous pouvons ensuite classer ce mouvement en zones cohérentes en utilisant un apprentissage non supervisé et produire des métriques directement utilisables pour diverses applications. Nous illustrons et validons cette même méthode sur l’analyse du flux sanguin chez l’embryon de poisson.

Continue reading

Segmentation d’IRM de cerveaux de nouveau-nés en quelques secondes à l’aide d’un réseau de neurones convolutif <i>pseudo-3D</i> et de transfert d’apprentissage

Abstract

L’imagerie par résonance magnétique (IRM) du cerveau est utilisée sur les nouveau-nés pour évaluer l’évolution du cerveau et diagnostiquer des maladies neurologiques. Ces examens nécessitent souvent une analyse quantitative des différents tissus du cerveau, de sorte qu’avoir une segmentation précise est essentiel. Dans cet article, nous proposons une méthode automatique rapide de segmentation en différents tissus des images IRM 3D de cerveaux de nouveau-nés ; elle utilise un réseau de neurones totalement convolutif (FCN) et du transfert d’apprentissage. Par rapport aux approches similaires qui reposent soit sur des patchs 2D ou 3D, soit sur des FCN totalement 3D, notre méthode est beaucoup plus rapide : elle ne prend que quelques secondes, et une seule modalité (T2) est nécessaire. Afin de prendre les informations 3D en compte, trois coupes 2D successives sont empilées pour former une image 2D en couleurs, dont l’ensemble sur tout le volume sert d’entrée à un FCN, pré-entraîné sur ImageNet pour la classification d’images naturelles. Nos expériences sur un ensemble de données de référence montrent que notre méthode obtient des résultats du niveau de ceux de l’état de l’art.

Continue reading