Domain Adaptation in remote sensing: increasing the portability of land-cover classifiers

Among the types of remote sensing acquisitions, optical images are certainly one of the most widely relied upon data sources for Earth observation. They provide detailed measurements of the electromagnetic radiation reflected or emitted by each pixel in the scene. Through a process termed supervised land-cover classification, this allows to automatically yet accurately distinguish objects at the surface of our planet. In this respect, when producing a land-cover map of the surveyed area, the availability of training examples representative of each thematic class is crucial for the success of the classification procedure. However, in real applications, due to several constraints on the sample collection process, labeled pixels are usually scarce. When analyzing an image for which those key samples are unavailable, a viable solution consists in resorting to the ground truth data of other previously acquired images. This option is attractive but several factors such as atmospheric, ground and acquisition conditions can cause radiometric differences between the images, hindering therefore the transfer of knowledge from one image to another. The goal of this Thesis is to supply remote sensing image analysts with suitable processing techniques to ensure a robust portability of the classification models across different images. The ultimate purpose is to map the land-cover classes over large spatial and temporal extents with minimal ground information. To overcome, or simply quantify, the observed shifts in the statistical distribution of the spectra of the materials, we study four approaches issued from the field of machine learning. First, we propose a strategy to intelligently sample the image of interest to collect the labels only in correspondence of the most useful pixels. This iterative routine is based on a constant evaluation of the pertinence to the new image of the initial training data actually belonging to a different image. Second, an approach to reduce the radiometric differences among the images by projecting the respective pixels in a common new data space is presented. We analyze a kernel-based feature extraction framework suited for such problems, showing that, after this relative normalization, the cross-image generalization abilities of a classifier are highly increased. Third, we test a new data-driven measure of distance between probability distributions to assess the distortions caused by differences in the acquisition geometry affecting series of multi-angle images. Also, we gauge the portability of classification models through the sequences. In both exercises, the efficacy of classic physically- and statistically-based normalization methods is discussed. Finally, we explore a new family of approaches based on sparse representations of the samples to reciprocally convert the data space of two images. The projection function bridging the images allows a synthesis of new pixels with more similar characteristics ultimately facilitating the land-cover mapping across images.

Effect of added dead space on sleep disordered breathing at high altitude.

OBJECTIVE: Sleep disordered breathing with central apnea or hypopnea frequently occurs at high altitude and is thought to be caused by a decrease in blood CO(2) level. The aim of this study was to assess the effects of added respiratory dead space on sleep disordered breathing.¦METHODS: Full polysomnographies were performed on 12 unacclimatized swiss mountaineers (11 males, 1 female, mean age 39±12 y.o.) in Leh, Ladakh (3500m). In random order, half of the night was spent with a 500ml increase in dead space through a custom designed full face mask and the other half without it.¦RESULTS: Baseline data revealed two clearly distinct groups: one with severe sleep disordered breathing (n=5, AHI>30) and the other with moderate to no disordered breathing (n=7, AHI<30). DS markedly improved breathing in the first group (baseline vs DS): apnea hypopnea index (AHI) 70.3±25.8 vs 29.4±6.9 (p=0.013), oxygen desaturation index (ODI): 72.9±24.1/h vs 42.5±14.4 (p=0.031), whereas it had no significant effect in the second group or in the total population. Respiratory events were almost exclusively central apnea or hypopnea. Microarousal index, sleep efficiency, and sleep architecture remained unchanged with DS. A minor increase in mean PtcCO(2) (n=3) was observed with DS.¦CONCLUSION: A 500ml increase in dead space through a fitted mask may improve nocturnal breathing in mountaineers with severe altitude-induced sleep disordered breathing.

Potencialidad de la modelización state-space y el filtro de Kalman para el anàlisis regional : una aplicación para el índice de actividad industrial

El déficit existente a nuestro país con respecto a la disponibilidad de indicadores cuantitativos con los que llevar a término un análisis coyuntural de la actividad industrial regional ha abierto un debate centrado en el estudio de cuál es la metodología más adecuada para elaborar indicadores de estas características. Dentro de este marco, en este trabajo se presentan las principales conclusiones obtenidas en anteriores estudios (Clar, et. al., 1997a, 1997b y 1998) sobre la idoneidad de extender las metodologías que actualmente se están aplicando a las regiones españolas para elaborar indicadores de la actividad industrial mediante métodos indirectos. Estas conclusiones llevan a plantear una estrategia distinta a las que actualmente se vienen aplicando. En concreto, se propone (siguiendo a Israilevich y Kuttner, 1993) un modelo de variables latentes para estimar el indicador de la producción industrial regional. Este tipo de modelo puede especificarse en términos de un modelo statespace y estimarse mediante el filtro de Kalman. Para validar la metodología propuesta se estiman unos indicadores de acuerdo con ella para tres de las cuatro regiones españolas que disponen d¿un Índice de Producción Industrial (IPI) elaborado mediante el método directo (Andalucía, Asturias y el País Vasco) y se comparan con los IPIs publicados (oficiales). Los resultados obtenidos muestran el buen comportamiento de l¿estrategia propuesta, abriendo así una línea de trabajo con la que subsanar el déficit al que se hacía referencia anteriormente

Voir le monde à travers les lunettes du métacontraste : un modèle mathématique inspiré de la psychologie sociale

Résumé La théorie de l'autocatégorisation est une théorie de psychologie sociale qui porte sur la relation entre l'individu et le groupe. Elle explique le comportement de groupe par la conception de soi et des autres en tant que membres de catégories sociales, et par l'attribution aux individus des caractéristiques prototypiques de ces catégories. Il s'agit donc d'une théorie de l'individu qui est censée expliquer des phénomènes collectifs. Les situations dans lesquelles un grand nombre d'individus interagissent de manière non triviale génèrent typiquement des comportements collectifs complexes qui sont difficiles à prévoir sur la base des comportements individuels. La simulation informatique de tels systèmes est un moyen fiable d'explorer de manière systématique la dynamique du comportement collectif en fonction des spécifications individuelles. Dans cette thèse, nous présentons un modèle formel d'une partie de la théorie de l'autocatégorisation appelée principe du métacontraste. À partir de la distribution d'un ensemble d'individus sur une ou plusieurs dimensions comparatives, le modèle génère les catégories et les prototypes associés. Nous montrons que le modèle se comporte de manière cohérente par rapport à la théorie et est capable de répliquer des données expérimentales concernant divers phénomènes de groupe, dont par exemple la polarisation. De plus, il permet de décrire systématiquement les prédictions de la théorie dont il dérive, notamment dans des situations nouvelles. Au niveau collectif, plusieurs dynamiques peuvent être observées, dont la convergence vers le consensus, vers une fragmentation ou vers l'émergence d'attitudes extrêmes. Nous étudions également l'effet du réseau social sur la dynamique et montrons qu'à l'exception de la vitesse de convergence, qui augmente lorsque les distances moyennes du réseau diminuent, les types de convergences dépendent peu du réseau choisi. Nous constatons d'autre part que les individus qui se situent à la frontière des groupes (dans le réseau social ou spatialement) ont une influence déterminante sur l'issue de la dynamique. Le modèle peut par ailleurs être utilisé comme un algorithme de classification automatique. Il identifie des prototypes autour desquels sont construits des groupes. Les prototypes sont positionnés de sorte à accentuer les caractéristiques typiques des groupes, et ne sont pas forcément centraux. Enfin, si l'on considère l'ensemble des pixels d'une image comme des individus dans un espace de couleur tridimensionnel, le modèle fournit un filtre qui permet d'atténuer du bruit, d'aider à la détection d'objets et de simuler des biais de perception comme l'induction chromatique. Abstract Self-categorization theory is a social psychology theory dealing with the relation between the individual and the group. It explains group behaviour through self- and others' conception as members of social categories, and through the attribution of the proto-typical categories' characteristics to the individuals. Hence, it is a theory of the individual that intends to explain collective phenomena. Situations involving a large number of non-trivially interacting individuals typically generate complex collective behaviours, which are difficult to anticipate on the basis of individual behaviour. Computer simulation of such systems is a reliable way of systematically exploring the dynamics of the collective behaviour depending on individual specifications. In this thesis, we present a formal model of a part of self-categorization theory named metacontrast principle. Given the distribution of a set of individuals on one or several comparison dimensions, the model generates categories and their associated prototypes. We show that the model behaves coherently with respect to the theory and is able to replicate experimental data concerning various group phenomena, for example polarization. Moreover, it allows to systematically describe the predictions of the theory from which it is derived, specially in unencountered situations. At the collective level, several dynamics can be observed, among which convergence towards consensus, towards frag-mentation or towards the emergence of extreme attitudes. We also study the effect of the social network on the dynamics and show that, except for the convergence speed which raises as the mean distances on the network decrease, the observed convergence types do not depend much on the chosen network. We further note that individuals located at the border of the groups (whether in the social network or spatially) have a decisive influence on the dynamics' issue. In addition, the model can be used as an automatic classification algorithm. It identifies prototypes around which groups are built. Prototypes are positioned such as to accentuate groups' typical characteristics and are not necessarily central. Finally, if we consider the set of pixels of an image as individuals in a three-dimensional color space, the model provides a filter that allows to lessen noise, to help detecting objects and to simulate perception biases such as chromatic induction.

Variational analysis of the Gross-Neveu model in an S^1 space

The Gross-Neveu model in an S^1 space is analyzed by means of a variational technique: the Gaussian effective potential. By making the proper connection with previous exact results at finite temperature, we show that this technique is able to describe the phase transition occurring in this model. We also make some remarks about the appropriate treatment of Grassmann variables in variational approaches.

The immune space: a concept and template for rationalizing vaccine development.

Abstract Empirical testing of candidate vaccines has led to the successful development of a number of lifesaving vaccines. The advent of new tools to manipulate antigens and new methods and vectors for vaccine delivery has led to a veritable explosion of potential vaccine designs. As a result, selection of candidate vaccines suitable for large-scale efficacy testing has become more challenging. This is especially true for diseases such as dengue, HIV, and tuberculosis where there is no validated animal model or correlate of immune protection. Establishing guidelines for the selection of vaccine candidates for advanced testing has become a necessity. A number of factors could be considered in making these decisions, including, for example, safety in animal and human studies, immune profile, protection in animal studies, production processes with product quality and stability, availability of resources, and estimated cost of goods. The "immune space template" proposed here provides a standardized approach by which the quality, level, and durability of immune responses elicited in early human trials by a candidate vaccine can be described. The immune response profile will demonstrate if and how the candidate is unique relative to other candidates, especially those that have preceded it into efficacy testing and, thus, what new information concerning potential immune correlates could be learned from an efficacy trial. A thorough characterization of immune responses should also provide insight into a developer's rationale for the vaccine's proposed mechanism of action. HIV vaccine researchers plan to include this general approach in up-selecting candidates for the next large efficacy trial. This "immune space" approach may also be applicable to other vaccine development endeavors where correlates of vaccine-induced immune protection remain unknown.

Pairing in a two-component ultracold Fermi gas: Phases with broken-space symmetries

We explore the phase diagram of a two-component ultracold atomic Fermi gas interacting with zero-range forces in the limit of weak coupling. We focus on the dependence of the pairing gap and the free energy on the variations in the number densities of the two species while the total density of the system is held fixed. As the density asymmetry is increased, the system exhibits a transition from a homogenous Bardeen-Cooper-Schrieffer (BCS) phase to phases with spontaneously broken global space symmetries. One such realization is the deformed Fermi surface superfluidity (DFS) which exploits the possibility of deforming the Fermi surfaces of the species into ellipsoidal form at zero total momentum of Cooper pairs. The critical asymmetries at which the transition from DFS to the unpaired state occurs are larger than those for the BCS phase. In this precritical region the DFS phase lowers the pairing energy of the asymmetric BCS state. We compare quantitatively the DFS phase to another realization of superconducting phases with broken translational symmetry: the single-plane-wave Larkin-Ovchinnikov-Fulde-Ferrell phase, which is characterized by a nonvanishing center-of-mass momentum of the Cooper pairs. The possibility of the detection of the DFS phase in the time-of-flight experiments is discussed and quantified for the case of 6Li atoms trapped in two different hyperfine states.