The paper introduces two resources for NLP, available with a GPL license: a dictionary of French compound words and a NooJ grammar which specifies a subset of compound patterns. Mots-clés : open source, ressources, dictionnaire, grammaire, mots composés

A vector segmentation of multi-spectral MSG2 images, based on the Markov formalism, is presented in this paper. The use of Markov fields take into account space and spectral interactions between the pixels of multi-component image. The use of these models requires an initial choice of their parameters obtained by algorithms of segmentation which can generate additional processing times. To avoi...

MOTS-CLÉS : micro-processeur, architectures reconfigurables, circuit FPGA, reconfiguration, reconfigurable computing, système reconfigurable

We consider the implementation of 16-bit floating point instructions on a Pentium 4 and a PowerPC G5 for image and media processing. By measuring the execution time of benchmarks with these new simulated instructions, we show that significant speed-ups are obtained compared to 32-bit FP versions. For image processing, the speed-up both comes from doubling the number of operations per SIMD instr...

Countryside automatic cartography is a real and decisive challenge in the current societal context (e.g., the famine risk analysis, the deforestation consequence analysis). In this paper, we propose a preliminary approach allowing the landscape characterization. More precisely, an itemset-based-technique is developed to extract crop types features. One of the main strengths of the proposed meth...

Image processing is often necessary for extracting the content of ancient documents. We present here techniques for restoring images and removing noise, extracting document structures (separating graphical elements and illustrations from text, extracting text lines) and, when possible, recognizing the textual or musical symbols which may be present in the image. These techniques, which are clas...

Frequently, medical queries are labeled with the best suitable image retrieval model. This manual classification is done by domain experts. In this paper, we propose a new approach to automatically learn how to classify medical queries into a set of retrieval models. First, we generate a set of class association rules which combine query features with image retrieval models. Then, we select the...

This paper presents a new area in the domain of Temporal Information Retrieval, which has received much attention for the last few years. Temporality can be viewed as a way of diversifying Web search results in response to any given user’s query. In this paper, we describe a study about image timestamping for temporal web image retrieval. This process imposes a large amount of theoretical and t...

— MOTS-CLÉS : Découverte d’abstractions, Analyse Formelle de Concepts, Analyse Relationnelle de Concepts, ingénierie des modèles, diagrammes de classes, composants, architectures

Terminological acquisition for identifying keywords of scientific articles The challenge DEFT2012 aims at automatically identifying the keywords chosen by the authors of scientific articles in the Humanities. A keyword list is provided within the track 1. We propose to exploit terminological acquisition approaches. The extracted terms are also sorted and filtered according to their position in ...