Explore the vast range of titles available.

Spatial autocorrelation is inherent to remotely sensed data. Nearby pixels are more similar than distant ones. This property can help to improve the classification performance, by adding spatial or contextual features into the model. However, it can also lead to overestimation of generalisation capabilities, if the spatial dependence between training and test sets is ignored. In this paper, we review existing approaches that deal with spatial autocorrelation for image classification in remote sensing and demonstrate the importance of bias in accuracy metrics when spatial independence between the training and test sets is not respected. We compare three spatial and non-spatial cross-validation strategies at pixel and object levels and study how performances vary at different sample sizes. Experiments based on Sentinel-2 data for mapping two simple forest classes show that spatial leave-one-out cross-validation is the better strategy to provide unbiased estimates of predictive error. Its performance metrics are consistent with the real quality of the resulting map contrary to traditional non-spatial cross-validation that overestimates accuracy. This highlight the need to change practices in classification accuracy assessment. To encourage it we developped Museo ToolBox, an open-source python library that makes spatial cross-validation possible.

The free to use Sentinel-2 (S2) sensors with 5-day revisit time at high spatial resolution in 10 spectral bands is a revolution in the remote sensing domain. Including 6 spectral bands in the near infrared, with 3 dedicated for the red-edge (where the vegetation significatively increases), these european satellites are very promising for mapping tree species distribution at a national scale. Here, we study the contribution of three one-year S2 Satellite Image Time Series (SITS) for mapping deciduous species distribution in the southwest of France. The annual cycle of vegetation (called phenology) can contribute to the identification of tree species. For some specific dates, species can have different phenological behaviours (senesence, flowering…). To train and validate the maps, we used the Support Vector Machine algorithm with a spatial cross-validation method. To train the algorithm with the same number of samples per species, we decided to undersample each class to the smallest class using a K-means clustering method. Moreover, a Sequential Feature Selection (SFS) has been implemented to detect the optimal dates per species. Our results are promising with high accuracy for Red oak andWillow (average score of the three one-year respectively F1 = 0.99, F1 = 0.94) based on the optimal dates. However, it appears that the performances when using the each full SITS are far below the optimal dates models (average ΔF1 = 0.32). We did not find, except for Willow and Red oak, that the optimal dates were the same for each year. Perspectives is to find an algorithm robust to temporal or spectral noise and to smooth the time series.

Context Hoverflies are often used as bio-indicators for ecosystem conservation, but only few studies have actually investigated the key factors explaining their richness in woodlands. Objectives In a fragmented landscape in southwest France, we investigated the joint effects of woodland area, structural heterogeneity, connectivity and history on the species richness of forest-specialist hoverflies, and whether there was a time lag in the response of hoverflies to habitat changes, and tested the effect of spatiotemporal changes. Methods Current species richness was sampled in 48 woodlands using 99 Malaise traps. Structural variables were derived from a rapid habitat assessment protocol. Old maps and aerial photographs were used to extract past and present spatial patterns of the woodlands since 1850. Relationships between species richness and explanatory variables were explored using generalized linear models. Results We show that current habitat area, connectivity, historical continuity and the average density of tree-microhabitats explained 35 % of variation in species richness. Species richness was affected differently by changes in patch area between 1979 and 2010, depending on woodland connectivity. In isolated woodlands, extinction debt and colonization credit were revealed, showing that even several decades are not sufficient for hoverflies to adapt to landscape-scale habitat conditions. Conclusions These findings emphasise the importance of maintaining connectedness between woodlands, which facilitates the dispersion in a changing landscape. Our results also highlight the benefits of using a change-oriented approach to explain the current distribution patterns of species, especially when several spatial processes act jointly.

The current context of availability of Earth Observation satellite data at high spatial and temporal resolutions makes it possible to map large areas. Although supervised classification is the most widely adopted approach, its performance is highly dependent on the availability and the quality of training data. However, gathering samples from field surveys or through photo interpretation is often expensive and time-consuming especially when the area to be classified is large. In this paper we propose the use of an active learning-based technique to address this issue by reducing the labelling effort required for supervised classification while increasing the generalisation capabilities of the classifier across space. Experiments were conducted to identify poplar plantations in three different sites in France using Sentinel-2 time series. In order to characterise the age of the identified poplar stands, temporal means of Sentinel-1 backscatter coefficients were computed. The results are promising and show the good capacities of the active learning-based approach to achieve similar performance (Poplar F-score  90%) to traditional passive learning (i.e. with random selection of samples) with up to 50% fewer training samples. Sentinel-1 annual means have demonstrated their potential to differentiate two stand ages with an overall accuracy of 83% regardless of the cultivar considered.

L’article dresse un portrait de la place qu’occupe Simenon dans les manuels scolaires italiens et tente de voir si et de quelle manière ses romans sont exploités en classe de langue. Passant en revue différents manuels de littérature destinés principalement aux lycéens, l’auteur tente de voir de quelle façon est appréhendé le romancier belge et si l’on remarque des constantes dans le choix des romans et extraits présentés dans les anthologies publiées dans les maisons d’édition italiennes. Il se penche également brièvement sur le phénomène des livres audio, versions simplifiées de grandes œuvres du patrimoine littéraire français. Il en résulte que Simenon est malheureusement trop souvent négligé et que seule sa production « policière » est envisagée. Perçu uniquement comme un auteur de polars, il est probablement parfois simplifié à outrance à des fins didactiques, étant davantage exploité dans un but linguistique que pour ses qualités littéraires dans l’enseignement du FLE. This article gives an idea of the importance Simenon is given in Italian schoolbooks and tries to see whether and in which way his novels are exploited in language classes. Going through various literature manuals mainly meant for high school students, the author attempts to determine how this Belgian novelist is perceived and whether it is mainly the same novels and extracts that are introduced in anthologies published by Italian publishing houses. He also briefly examines the phenomenon of audiobooks which are simplified versions of the main works belonging to the French literature patrimonial. The conclusion is unfortunately that Simenon is all too often neglected and that only his detective production is considered. Solely perceived as an author of thrillers, he is more than likely sometimes excessively simplified for didactic reasons as he is most often exploited for linguistic purposes rather than for his literature qualities in the teaching of French as a foreign language.

Numéro de revue consacré aux différents liens entre Simenon et l'Italie, tant en ce qui concerne la place des Italiens dans ses romans et ses mémoires que sa réception dans la péninsule.

L'article a l'objectif de décrire brievement la position inconfortable du professeur de FLE originaire d'un pays francophone autre que la France, qui a du mal a trouver un équilibre entre la nécessité de se référer a une norme pédagogique préconisée par le systeme scolaire - celle de l'Hexagone - et ses propres pratiques linguistiques, son identité se traduisant en partie a travers son langage et son accent. Les méthodes de langue, bien qu'elles abordent succinctement la francophonie d'un point de vue culturel, ne laissent cependant aucune place a la variation linguistique d'un point de vue diatopique, contribuant par la meme a accentuer ce malaise, créant de fait un écart entre le vécu ethnoculturel et linguistique de l'enseignant francophone de langue maternelle et un idiome métropolitain qui lui est quelque peu \"étranger\" mais qui est celui qu'il devra pourtant transmettre aux apprenants.

Introduction au numéro de \"Francofonia\" sur Simenon et l'Itale