Éditeur(s) : HAL CCSD Springer Verlag (Germany)
Résumé : International audience
Surface displacements due to temporal changes in environmental mass redistributions are observable in the coordinate time series of many Global Navigation Satellite System (GNSS) sites. In this study, we investigated the effect of loading on estimates of tectonic velocity computed from campaign-style GNSS observations. The study region is in the Pyrenees mountain range between France and Spain (ResPyr campaigns). In this area, seismic activity is continuous and moderate and the expected amplitude of the horizontal tectonic velocity is less than 0.5 mm/yr. In order to determine the velocity, 4 sparse GNSS campaigns were carried out from 1995 to 2010. Considering this small rate of deformation, loading phenomena can contribute a non-negligible artifact to the velocity computation that could affect our geodynamical interpretation. In this investigation, we specifically considered the atmospheric, hydrological, and non-tidal ocean loading phenomena. The computed loading deformations for this region show the horizontal displacements are dominated by the non-tidal ocean loading (maximum 4 mm for the North and 3.1 mm for the East components); the main vertical contributions are due to the atmospheric and continental water storage loading (maximum 14.3 for the atmosphere and 8.1 mm for the hydrology, respectively). We have found that the dominant loading effect on the horizontal velocity is the non-tidal ocean loading (mean of 0.11 mm/yr), whereas the vertical component is dominated by the hydrological loading (mean of 0.21 mm/yr). Since the study area is in a mountainous region, we also analyzed the difference between the atmospheric and the topography dependent atmospheric loading models at our GNSS campaign sites. We did not find any significant difference between the two atmospheric loading models in terms of horizontal velocity. Finally, we performed simulations to identify the optimum timing and frequency of future GNSS campaigns in this area that would minimize the loading effects on tectonic velocity estimates.
ISSN: 0039-3169

hal-00963629
https://hal.archives-ouvertes.fr/hal-00963629
DOI : 10.1007/s11200-012-0458-2

Éditeur(s) : HAL CCSD
Résumé : International audience
“Affective Computing is computing that relates to, arises from, or deliberately influences emotions” (R.W. Picard, MIT Press, 1997). Specifically, psychosocial equilibrium is now considered as a major issue for both individual and public health. Predicting human emotions via non-intrusive methods is a great challenge triggered by the rise of intelligent assisting systems. In order to determine the emotional state of a subject having a potential psychosocial disorder, it is required to process both neutral and expressive emotions. A computer-engineered method aiming to provide emotional information from videos is proposed. An original approach for affect classification and its interpretation is presented.The database is part of the HUMAINE database, and consists of a large volume of video clips, each displaying individuals, depressed or not, talking with a conversational agent. Both the high-dimensionality of data and its temporal aspect require a pre-training that consists of cleaning, sorting and feature selection, in order to make relevant information usable for further processing by MachineLearning methods. Next step consists in classifying the affective state of subjects according to their positivity or negativity (Russell's model, 1980). By means of six independent classifiers, each running a Support Vector Machine algorithm, separate performances are obtained for each class of emotions. The results reveal an interesting “neutral area phenomenon”, where emotions tend to be harder to be detected.Though more frequent, neutral emotions seem to carry less discriminant information than expressive ones. This issue will be a key for designing a tuning step in further research.
Caribbean Academy Of Science
Deshaies, France

hal-01408029
https://hal.univ-antilles.fr/hal-01408029

Éditeur(s) : HAL CCSD Seismological Society of America
Résumé : International audience
We present an improvement of the maximum intersection (MAXI) method in which absolute earthquake location is defined by the maximum intersection number of hyperbolic equal differential time (EDT) volumes (one EDT volume being described as all grid nodes satisfying the arrival‐time differences between two stations, ± a tolerance value known as TERR). This 3D technique is well adapted to a strongly heterogeneous environment, avoids the depth versus origin‐time trade‐off (even using P arrivals alone), and objectively filters possible erroneous arrival times. Improvements consist of introducing an iterative multiscale approach on the TERR parameter and conducting a hierarchy of grid discretization to refine the preliminary solution (multigrid algorithm). The outlier cleaning process is strengthened by conducting statistical analyses based on the number of EDT volume intersections per station, rather than on travel‐time residues. Processing two series of synthetic arrival times computed in the 3D seismological configuration of the southernmost Ryukyu subduction zone, offshore Taiwan, allows us to (1) confirm MAXI robustness, even when arrival times are highly disturbed or azimuthal coverage largely reduced, and (2) show that MAXI, by limiting trade‐offs between origin time and depth and between epicentral position and depth, is less affected by a wrong velocity model than a technique that minimizes travel‐time residues. To conclude, we apply the MAXI method to locate earthquakes recorded by the combined Central Weather Bureau of Taiwan (CWB) and the Japanese Meteorological Agency (JMA) networks (1991-2008) within a georealistic a priori 3D P‐velocity model. Hypocenter determinations are selected based on MAXI confidence factors, and earthquake clustering along well‐known tectonic features witnesses hypocenter determination accuracy.
ISSN: 0037-1106

hal-00725227
https://hal.archives-ouvertes.fr/hal-00725227
DOI : 10.1785/0120100311

Éditeur(s) : HAL CCSD
Résumé : National audience
Designing systems able to interact with students in a natural manner is a complex and far from solved problem. A key aspect of natural interaction is the ability to understand and appropriately respond to human emotions. This paper details our response to the continuous Audio/Visual Emotion Challenge (AVEC'12) whose goal is to predict four affective signals describing human emotions. The proposed method uses Fourier spectra to extract multi-scale dynamic descriptions of signals characterizing face appearance, head movements and voice. We perform a kernel regression with very few representative samples selected via a supervised weighted-distance-based clustering, that leads to a high generalization power. We also propose a particularly fast regressor-level fusion framework to merge systems based on different modalities. Experiments have proven the efficiency of each key point of the proposed method and our results on challenge data were the highest among 10 international research teams.
IHM'14, 26e conférence francophone sur l'Interaction Homme-Machine
Lille, France

hal-01089628
https://hal.archives-ouvertes.fr/hal-01089628
https://hal.archives-ouvertes.fr/hal-01089628/document
https://hal.archives-ouvertes.fr/hal-01089628/file/p44-nicole.pdf

Éditeur(s) : HAL CCSD
Résumé : International audience
Symbiosis is a widespread phenomenon in the biosphere. In marine benthic environments breakdown of symbiosis is responsible for coral bleaching and dramatically affects coral reef ecosystems functioning [1]. In the water column, planktonic organisms are key component of pelagic ecosystems and a number of species form mutualistic association with microalgae forming a photosymbiosis [2]. Here we intended to investigate the genetic basis of photosymbiosis, through a transcriptomic approach on marine plankton organisms. We focused more specifically on associations occurring between radiolarian host (protist, zooplankton) and dinoflagellates symbiont (protist, phytoplankton) living inside the host cell. It has recently been highlighted that these holobionts are widespread in oligotrophic open oceans and there are also evidences of their fundamental implication in biogeochemical carbon, silica and strontium cycles [3, 4, 5, 6]. RNA-seq technologies allow obtaining an unprecedented amount of data for unicellular organisms isolated from the environment [7]. The study of such non-model holobionts datasets requires de novo assembly, which implies considerable computational resources and the potential production of chimeric sequences [8, 9].We therefore developed an original strategy aiming at accelerating and improving de novo assembly for holobiont datasets. We chose SIMKA [10] a fast kmer-based method initially developed to estimate the similarity between numerous metagenomic datasets, and which has been recently adapted to extract their common sequences. As our symbionts are identified, we used SIMKA to compare our holobionts transcriptomes to publicly available dinoflagellates transcriptomes [11] generating two datasets, one composed of reads from the symbionts and another with reads from the host (for which no reference data are currently available). Independent assemblies were then performed in parallel, accelerating the study process, and minimizing the proportion of resulting chimeric sequences. Our strategy produced a unique and comprehensive genomic dataset for Radiolaria [12, 13], and offers a pragmatic, large scale, comparison strategy to assemble and study holobionts [9]. Our new sequences obtained from holobionts study will be used for phylogenomics investigation, as reference for environmental metagenomic studies and ultimately to understand and characterize the molecular basis of symbiotic relationships in the plankton.Bibliografic references :[1] Simon K. Davy, Denis Allemand, and Virginia M. Weis. “Cell Biology of Cnidarian-Dinoflagellate Symbiosis”. In: Microbiology and Molecular Biology Reviews 76.2 (June 1, 2012), pp. 229–261. issn: 1092-2172, 1098- 5557. doi: 10.1128/MMBR.05014-11.[2] Janouškovec, J. et al. “Major transitions in dinoflagellate evolution unveiled by phylotranscriptomics”. in: PNAS 201614842 (2016). doi:10.1073/pnas.1614842114[3] Tristan Biard et al. “In situ imaging reveals the biomass of giant protists in the global ocean”. In: Nature advance online publication (Apr. 20, 2016). issn: 0028-0836. doi: 10.1038/nature17652.[4] Lionel Guidi et al. “Plankton networks driving carbon export in the olig- otrophic ocean”. In: Nature 532.7600 (Apr. 28, 2016), pp. 465–470. issn: 0028-0836. doi: 10.1038/nature16942.[5] Diane K. Stoecker et al. “Acquired phototrophy in aquatic protists”. In: Aquatic Microbial Ecology 57.3 (Nov. 24, 2009), pp. 279–310. doi: 10.3354/ame01340.[6] Johan Decelle et al. “An original mode of symbiosis in open ocean plankton”. In: Proceedings of the National Academy of Sciences 109.44 (Oct. 30, 2012), pp. 18000–18005. issn: 0027-8424, 1091-6490. doi: 10. 1073/pnas.1212303109.[7] Sergio Balzano et al. “Transcriptome analyses to investigate symbiotic relationships between marine protists”. In: Frontiers in Microbiology 6 (Mar. 17, 2015). issn: 1664-302X. doi: 10.3389/fmicb.2015.00098.[8] Li, B. et al. “Evaluation of de novo transcriptome assemblies from RNA-Seq data”. In: Genome Biology 15, 553 (2014).[9] Sangwan, N., Xia, F. & Gilbert, J. A. “Recovering complete and draft population genomes from metagenome datasets”. In: Microbiome 4, 8 (2016).[10] Gaëtan Benoit et al. “Multiple comparative metagenomics using mul- tiset k -mer counting”. In: PeerJ Computer Science 2 (Nov. 14, 2016). doi: 10.7717/peerj-cs.94.[11] Patrick J. Keeling et al. “The Marine Microbial Eukaryote Transcriptome Sequencing Project (MMETSP): Illuminating the Functional Diversity of Eukaryotic Life in the Oceans through Transcriptome Sequencing”. In: PLOS Biol 12.6 (June 2014), e1001889. issn: 1545-7885. doi: 10.1371/journal.pbio.1001889.[12] Burki, F. et al. “Evolution of Rhizaria: new insights from phylogenomic analysis of uncultured protists”. In BMC Evolutionary Biology 10, 377 (2010).
International Conference on Holobionts
Paris, France

hal-01575069
https://hal.inria.fr/hal-01575069

Éditeur(s) : HAL CCSD Wiley
Résumé :
How Quaternary climatic and geological disturbances influenced the composition of Neotropical forests is hotly debated. Rain fall and temperature changes during and/or immediately after the Last Glacial Maximum (LGM) are thought to have strongly affected the geographical distribution and local abundance of tree species. The paucity of the fossil records in Neotropical forests prevents a direct reconstruction of such processes. To describe community-level historical trends in forest composition, we turned therefore to inferential methods based on the reconstruction of past demographic changes. In particular, we modelled the history of rainforests in the Eastern Guiana Shield over a time scale of several thousand generations, through the application of Approximate Bayesian Computation (ABC) and Maximum Likelihood (ML) methods to diversity data at nuclear and chloroplast loci in eight species or subspecies of rainforest trees. Depending on the species and on the method applied, we detected population contraction, expansion or stability, with a general trend in favour of stability or expansion, with changes presumably having occurred during or after the LGM. These findings suggest that Guiana Shield rain forests have globally persisted, while expanding, through the Quaternary, but that different species have experienced different demographic events, with a trend towards the increase in frequency of light-demanding, disturbance-associated species.
ISSN: 0962-1083

hal-01512150
https://hal.archives-ouvertes.fr/hal-01512150
https://hal.archives-ouvertes.fr/hal-01512150/document
https://hal.archives-ouvertes.fr/hal-01512150/file/Barthe_et_al-_%7B0B6A8C5E-DDFF-4C4E-822C-4E7BC22A6C9B%7D.pdf
DOI : 10.1111/mec.13949

Éditeur(s) : HAL CCSD
Résumé : The Restefond fault, located in the Late Eocene-Early Oligocene Alpine foreland basin, affects the well lithified and low porosity Grès d'Annot. The fault core zone is characterized by the occurrence of highly deformed sandstone lenses. Deformation inside the lenses corresponds to mm to sub-mm-spaced cleavage planes rich in phyllosilicates and up to cm-thick and dm-long quartz-calcite pure extensional veins. The cleavages are mostly composed of newly-formed synkinematic white mica and chlorite. By using thermodynamic thermometers based on the chemical composition of chlorite, a temperature of 200 ± 20 °C of fault activity was computed. This temperature shows that the Restefond fault was active at burial conditions comprised between 6.5 and 8 km, assuming a mean geothermal gradient between 25 and 30 °Ckm−1. The petrophysic properties of sandstones from the core zone and in the hanging and foot wall of the fault were determined on drilled plugs following three spacial directions. The permeability of the highly deformed sandstone from the core zone is about one order of magnitude higher than in the host rock. This increase in permeability occurs in the direction parallel to the S-C structures and is explained by the occurrence of well-connected micropores localized between platy phyllosilicates. This study shows that the fault petrophysic properties are mostly controlled by the precipitation of synkinematic phyllosilicates under deep burial conditions.
Swiss Journal of Geosciences

hal-00795551
https://hal.archives-ouvertes.fr/hal-00795551
DOI : 10.1007/s00015-012-0099-z

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
3D seismic data located in the Gjallar Ridge (Vøring Basin, offshore Norway) reveals a closely-spaced polygonal fault system affecting more than 800 m of homogeneous mud-dominated Quaternary and Tertiary sequences. As some faults reach the modern seafloor, they represent an active polygonal fault system at present day. Even if the processes remain unclear and are still under debate, it is generally agreed that the initiation of polygonal faults is the result of shallow burial dewatering of fine-grained unconsolidated sediments by volumetric compaction. 3D seismic data are commonly interpreted by propagating horizons automatically and by picking faults manually. However, in the case of polygonal fault intervals, this approach is time consuming due to the huge number of faults and because automatic propagation can be misleading. In this study, we applied a new technique of 3D seismic interpretation based on a sequential stratigraphy analysis, using the new PaleoScan© software (Eliis Company). It allowed us to build a 3D geological model computing more than 300 horizons within the faulted intervals. We then used the coherency attribute, depicting anomalies in the shape of seismic waveform like faults, in order to constrain a possible link between fault distribution and stratigraphic levels. Our approach allows fault throws to be calculated in milliseconds on any polygonal fault plane. The result shows that fault segments have been reactivated by dip-linkage. Distribution of faults depends on mechanical units, intervals characterized by different petrophysical properties, which are independent from lithological and diagenetic changes. According to these results, we propose a model showing the evolution of polygonal fault intervals in which faulting stages are separated by a quiescence phase during burial. A first tier of polygonal faults is initiated at a specific depth, according to the Cam-clay model. Then, following a period of quiescence during which mud-rich sediments continued to accumulate, new fault segments are initiated above the first mechanical unit and within this undeformed interval. New nucleated faults then connect downward to pre-existing underlying polygonal fault system, thus progressively increasing the thickness of the faulted interval.
ISSN: 0025-3227

hal-00772470
https://hal.archives-ouvertes.fr/hal-00772470
DOI : 10.1016/j.margeo.2012.07.016

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Deep-seated gravitational spreading (DSGS) affects the slopes of formerly glaciated mountain ridges. On Mars, DSGS has played a key role in shaping the landforms of the giant Valles Marineris troughs. Though less spectacular, DSGS is common in terrestrial orogens, where understanding its mechanics is critical in the light of the ongoing climate change because it is a potential source of catastrophic landslides in deglaciated valleys. We conducted parametric numerical studies in order to identify important factors responsible for DSGS initiation. DSGS models are computed using an elastoviscoplastic finite element code. Using ADELI's software, we reproduce topographic ridge spreading under the effect of valley unloading. Two types of spreading topographic ridges are investigated, homogeneous or with horizontal rheological layering. We find that gravitational instabilities are enhanced by high slopes, which increase gravitational stress, and low friction and cohesion, which decrease yield stress. In the unlayered ridge, instability is triggered by glacial unloading with plastic strain concentration inside the ridge and at the base of the high slopes. Vertical fractures develop in the upper part of the slope, potentially leading to fault scarps. Ridge homogeneity promotes a deformation mode controlled by uphill-facing normal faulting and basal bulging. In the second case, the ridge encompasses horizontal geological discontinuities that induce rock mass anisotropy. Discontinuity located at the base of the slope accumulates plastic strain, leading to the formation of a sliding plane evolving into a landslide. The presence of a weak layer at ridge base therefore promotes another slope deformation mode ending up with catastrophic failure. Mechanical conditions and slope height being equal, these conclusions can probably be extrapolated to Earth. Compared with Mars, DSGS on Earth is inhibited because terrestrial topographic gradients are lower than in Valles Marineris, an effect counterbalanced by increased gravitational stress, where the intensity of deformation is enhanced because of the Earth gravity potential.
ISSN: 0169-555X

hal-01420007
https://hal.archives-ouvertes.fr/hal-01420007
DOI : 10.1016/j.geomorph.2016.06.011

Éditeur(s) : HAL CCSD Springer Verlag
Résumé : International audience
In France, reducing social health inequalities has become an explicit goal of health policies over the past few years, one of its objectives is specifically the reduction of the perinatal mortality rate. This study investigates the association between infant mortality and social deprivation categories at a small area level in the Lille metropolitan area, in the north of France, to identify census blocks where public authorities should prioritize appropriate preventive actions. We used census data to establish a neighbourhood deprivation index whose multiple dimensions encompass socioeconomic characteristics. Infant mortality data were obtained from the Lille metropolitan area municipalities to estimate a death rate at the census tract level. We used Bayesian hierarchical models in order to reduce the extra variability when computing relative risks (RR) and to assess the associations between infant mortality and deprivation. Between 2000 and 2009, 668 cases of infant death occurred in the Lille metropolitan area (4.2 per 1,000 live births). The socioeconomic status is associated with infant mortality, with a clear gradient of risk from the most privileged census blocks to the most deprived ones (RR = 2.62, 95 % confidence interval [1.87; 3.70]). The latter have 24.6 % of families who were single parents and 29.9 % of unemployed people in the labor force versus 8.5 % and 7.7 % in the former. Our study reveals socio-spatial disparities in infant mortality in the Lille metropolitan area and highlights the census blocks most affected by the inequalities. Fine spatial analysis may help inform the design of preventive policies tailored to the characteristics of the local communities.
ISSN: 1092-7875

hal-00876255
https://hal.archives-ouvertes.fr/hal-00876255
DOI : 10.1007/s10995-013-1251-8
PUBMED : 23479336

Éditeur(s) : HAL CCSD
Résumé : International audience
The large majority of methods proposed in literature for handwriting recognition assume that words are produced drawing large parts of the ink without lifting the pen, other than horizontal bars and dots. This fundamental assumption, however, does not always hold: while some educational systems provide explicit training for producing continuous handwriting, minimizing the number of pen-up during the production of a word, others do not. As a consequence, whenever the handwriting presents pen-up within a word, the recognition performance can drop significantly. In a preliminary study, we presented an algorithm for discriminating among different types of ink appearing in handwriting, namely isolated characters, cursive, dots, horizontal and vertical bars, based on the use of a suitable set of features. In this paper, we have characterized the discriminative power of each considered feature according to different measures and we have proposed a method for combining the different feature rankings. We have also used the Fischer's Linear Discriminant Analysis (LDA) for exhaustively selecting the best feature subsets with increasing number of features. Finally, we have compared the results obtained by using the feature subsets provided by LDA with those obtained with the feature subsets selected according to our feature ranking. The experimental results, on different datasets of handwritten words, showed that our approach successfully achieves its aim allowing to reduce the computational cost without affecting the overall performance of the recognition process.
17th Biennial Conference of the International Graphonomics Society
Pointe-à-Pitre, Guadeloupe

hal-01165877
https://hal.univ-antilles.fr/hal-01165877
https://hal.univ-antilles.fr/hal-01165877/document
https://hal.univ-antilles.fr/hal-01165877/file/IGS_2015_submission_13.pdf