Le développement de systèmes de mesures et de suivi non-invasifs, fiables, robustes et utilisablesen autonomie est de première importance pour le confort et l’implication de la personne prise en charge dansson parcours de santé. En collaboration avec la société IDS SA, ce travail a été initié et motivé par la volontéde développer des méthodes de traitement et d’analyse dédiées à l’exploration fonctionnelle de mesures phy-siologiques non-invasives d’une part, et le développement de solutions pour la santé et l’autonomie, d’autre part.Les aspects fondamentaux de ce travail doctoral ont pour objectifs la mise au point et la validation de méthodesde décomposition modale pour : (1) l’extraction et la reconstruction des composantes cardiaques d’un signalpléthysmographique et (2) l’estimation d’un indice de complexité témoignant de modifications de stratégiesde contrôle postural. Il est attendu que ces méthodes soient : (1) locales pour gérer les non-stationnarités, (2)lisibles pour permettre l’identification des composantes et (3) adaptatives au sens d’un paramétrage a minima.`A cette fin, la connaissance du contexte physiologique et des modèles spectraux attenants sont employés toutautant comme un guide dans le choix et l’utilisation qu’une grille de lecture dans l’exploitation des méthodesengagées.Dans la première partie de ce travail, à travers le développement d’un banc de test sur signaux simulés, nousavons démontré la supériorité d’une variante de la décomposition modale empirique en comparaison à sa formeoriginale. Par la suite, nous avons établi l’utilisabilité, avec des hypothèses a minima, d’une représentationmodale récente, la transformée par Synchrosqueezing (SQT).Dans la seconde partie de ce travail, nous avons mis au point un indice de complexité en échelles basé sur la SQTet dont le fenêtrage temporel est entièrement déterminé par la dynamique spectrale de la représentation. Aupréalable, nous avons mis en place un protocole expérimental pour identifier les limites des méthodes existanteset légitimer une approche spectrale de l’entropie. Notre indice a été validé sur signaux simulés et testé sur si-gnaux réels où il a pu mettre en évidence le phénomène de repondération sensorielle à la suite d’une perturbation.Les aspects appliqués de ce travail doctoral, quant à eux, s’articulent autour du développement de solutionsdédiées : (1) à la prévention du risque de chute et (2) au suivi des activités de la vie quotidienne.Dans cette dernière partie, nous nous sommes intéressés : (1) à l’évaluation des troubles latéralisés de l’équilibre,(2) à l’évaluation de la prise en charge en charge de troubles posturaux, (3) à la conception et la validationd’un outil de mesure et de rééducation des capacités de contrôle postural intégré dans un smartphone, et (4) àla détection de dérives comportementales par le développement d’un indice de persévération et son applicationau suivi du rythme nycthéméral des activités de la vie quotidienne. / L'auteur n'a pas fourni de résumé en anglais

Scale-invariant interest points have found several highly successful applications in computer vision, in particular for image-based matching and recognition. This paper presents a theoretical analysis of the scale selection properties of a generalized framework for detecting interest points from scale-space features presented in Lindeberg (Int. J. Comput. Vis. 2010, under revision) and comprising: an enriched set of differential interest operators at a fixed scale including the Laplacian operator, the determinant of the Hessian, the new Hessian feature strength measures I and II and the rescaled level curve curvature operator, as well as an enriched set of scale selection mechanisms including scale selection based on local extrema over scale, complementary post-smoothing after the computation of non-linear differential invariants and scale selection based on weighted averaging of scale values along feature trajectories over scale. A theoretical analysis of the sensitivity to affine image deformations is presented, and it is shown that the scale estimates obtained from the determinant of the Hessian operator are affine covariant for an anisotropic Gaussian blob model. Among the other purely second-order operators, the Hessian feature strength measure I has the lowest sensitivity to non-uniform scaling transformations, followed by the Laplacian operator and the Hessian feature strength measure II. The predictions from this theoretical analysis agree with experimental results of the repeatability properties of the different interest point detectors under affine and perspective transformations of real image data. A number of less complete results are derived for the level curve curvature operator. / <p>QC 20121003</p> / Image descriptors and scale-space theory for spatial and spatio-temporal recognition

This thesis, within the subfield of computer science known as computer vision, deals with the use of scale-space analysis in early low-level processing of visual information. The main contributions comprise the following five subjects: The formulation of a scale-space theory for discrete signals. Previously, the scale-space concept has been expressed for continuous signals only. We propose that the canonical way to construct a scale-space for discrete signals is by convolution with a kernel called the discrete analogue of the Gaussian kernel, or equivalently by solving a semi-discretized version of the diffusion equation. Both the one-dimensional and two-dimensional cases are covered. An extensive analysis of discrete smoothing kernels is carried out for one-dimensional signals and the discrete scale-space properties of the most common discretizations to the continuous theory are analysed. A representation, called the scale-space primal sketch, which gives a formal description of the hierarchical relations between structures at different levels of scale. It is aimed at making information in the scale-space representation explicit. We give a theory for its construction and an algorithm for computing it. A theory for extracting significant image structures and determining the scales of these structures from this representation in a solely bottom-up data-driven way. Examples demonstrating how such qualitative information extracted from the scale-space primal sketch can be used for guiding and simplifying other early visual processes. Applications are given to edge detection, histogram analysis and classification based on local features. Among other possible applications one can mention perceptual grouping, texture analysis, stereo matching, model matching and motion. A detailed theoretical analysis of the evolution properties of critical points and blobs in scale-space, comprising drift velocity estimates under scale-space smoothing, a classification of the possible types of generic events at bifurcation situations and estimates of how the number of local extrema in a signal can be expected to decrease as function of the scale parameter. For two-dimensional signals the generic bifurcation events are annihilations and creations of extremum-saddle point pairs. Interpreted in terms of blobs, these transitions correspond to annihilations, merges, splits and creations. Experiments on different types of real imagery demonstrate that the proposed theory gives perceptually intuitive results. / <p>QC 20120119</p>

Autonomous driving technology has the potential to revolutionize the way we think about transportation and its impact on society. Perceiving the environment is a key aspect of autonomous driving, which involves multiple computer vision tasks. Multi-scale deep learning has dramatically improved the performance on many computer vision tasks, but its practical use in autonomous driving is limited by the available resources in embedded systems. Multi-task learning offers a solution to this problem by allowing more compact deep learning models that share parameters between tasks. However, not all tasks benefit from being learned together. One way of avoiding task interference during training is to learn tasks in sequence, with each task providing useful information for the next – a scheme which builds on transfer learning. Multi-task and transfer dynamics are both concerned with the relationships between tasks, but have previously only been studied separately. This Master’s thesis investigates how different computer vision tasks relate to each other in the context of multi-task and transfer learning, using a state-ofthe-art efficient multi-scale deep learning model. Through an experimental research methodology, the performance on semantic segmentation, depth estimation, and object detection were evaluated on the Virtual KITTI 2 dataset in a multi-task and transfer learning setting. In addition, transfer learning with a frozen encoder was compared to constrained encoder fine tuning, to uncover the effects of fine-tuning on task dynamics. The results suggest that findings from previous work regarding semantic segmentation and depth estimation in multi-task learning generalize to multi-scale learning on autonomous driving data. Further, no statistically significant correlation was found between multitask learning dynamics and transfer learning dynamics. An analysis of the results from transfer learning indicate that some tasks might be more sensitive to fine-tuning than others, suggesting that transferring with a frozen encoder only captures a subset of the complexities involved in transfer relationships. Regarding object detection, it is observed to negatively impact the performance on other tasks during multi-task learning, but might be a valuable task to transfer from due to lower annotation costs. Possible avenues for future work include applying the used methodology to real-world datasets and exploring ways of utilizing the presented findings for more efficient perception algorithms. / Självkörande teknik har potential att revolutionera transport och dess påverkan på samhället. Självkörning medför ett flertal uppgifter inom datorseende, som bäst löses med djupa neurala nätverk som lär sig att tolka bilder på flera olika skalor. Begränsningar i mobil hårdvara kräver dock att tekniker som multiuppgifts- och sekventiell inlärning används för att minska neurala nätverkets fotavtryck, där sekventiell inlärning bygger på överföringsinlärning. Dynamiken bakom både multiuppgiftsinlärning och överföringsinlärning kan till stor del krediteras relationen mellan olika uppdrag. Tidigare studier har dock bara undersökt dessa dynamiker var för sig. Detta examensarbete undersöker relationen mellan olika uppdrag inom datorseende från perspektivet av både multiuppgifts- och överföringsinlärning. En experimentell forskningsmetodik användes för att jämföra och undersöka tre uppgifter inom datorseende på datasetet Virtual KITTI 2. Resultaten stärker tidigare forskning och föreslår att tidigare fynd kan generaliseras till flerskaliga nätverk och data för självkörning. Resultaten visar inte på någon signifikant korrelation mellan multiuppgift- och överföringsdynamik. Slutligen antyder resultaten att vissa uppgiftspar ställer högre krav än andra på att nätverket anpassas efter överföring.