Descripteur
Documents disponibles dans cette catégorie (9301)
![](./images/expand_all.gif)
![](./images/collapse_all.gif)
Etendre la recherche sur niveau(x) vers le bas
Titre : Auxiliary tasks for the conditioning of generative adversarial networks Type de document : Thèse/HDR Auteurs : Cyprien Ruffino, Auteur ; Gilles Gasso, Directeur de thèse Editeur : Rouen [France] : Institut National des Sciences Appliquées INSA Rouen Année de publication : 2021 Importance : 136 p. Format : 21 x 30 cm Note générale : bibliographie
Pour obtenir le grade de Docteur de Normandie Université, Spécialité InformatiqueLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] classification du maximum a posteriori
[Termes IGN] classification non dirigée
[Termes IGN] classification par réseau neuronal convolutif
[Termes IGN] détection d'objet
[Termes IGN] reconstruction d'image
[Termes IGN] réseau antagoniste génératif
[Termes IGN] restauration d'imageIndex. décimale : THESE Thèses et HDR Résumé : (auteur) During the last decade, Generative Adversarial Networks (GANs) have caused a tremendous leap forward in image generation as a whole. Their ability to learn very complex, high-dimension distributions not only had a huge impact on the field of generative modeling, their influence extended to the general public at large. By being the first models able generate high-dimension photo-realistic images, GANs very quickly gained popularity as an image generation and photo manipulation technique. For example, their use as "filters" became common practice on social media, but they also allowed for the rise of Deepfakes, images that have been manipulated in order to fake the identity of a person. In this thesis, we explore the conditioning of Generative Adversarial Networks, that is influencing the generation process in order to control the content of a generated image. We focus on conditioning through auxiliary tasks, that is we explicitly implement additional objective to the generative model to complement the initial goal of learning the data distribution. First, we introduce generative modeling through several examples, and present the Generative Adversarial Networks framework. We discuss theoretical interpretations of GANs as well as its most prominent issues, notably the lack of stability during training of the model and the difficulty to generate diverse samples. We review classical techniques for conditioning GANs and propose an overview of recent approaches aiming to both solve the aforementioned issues and enhance the visual quality of the generated images. Afterwards, we focus on a specific generation task that requires conditioning : image reconstruction. In a nutshell, the problem consists in recovering an image from which we only have a handful of pixels available, usually around 0.5%. It stems from an application in geostatistics, namely the reconstruction of underground terrain from a reduced amount of expensive and difficult to obtain measurements. To do so, we propose to introduce an explicit auxiliary reconstruction task to the GAN framework which, in addition to a diversity-restoring technique, allows for the generation of high-quality images that respect the given measurements. Finally, we investigate a task of domain-transfer with generative models, specifically transferring images from the RGB color domain to the polarimetric domain. Polarimetric images bear hard constraints that directly stem from the physics of polarimetry. Leveraging on the cyclic-consistency paradigm, we extend the training of generative models with auxiliary tasks that push the generator towards enforcing the polarimetric constraints. We highlight that the approach manages to generate physically realistic polarimetric. Note de contenu : Introduction
1- Introduction to Generative Adversarial Networks
2- Image reconstruction as an auxiliary task to generative modeling
3- Domain-transfer with with auxiliary tasks for generative modeling
4- Conclusion and PerspectivesNuméro de notice : 28640 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse française Note de thèse : Thèse de Doctorat : Informatique : Normandie : 2021 Organisme de stage : LITIS DOI : sans En ligne : https://tel.hal.science/tel-03517304/ Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99721 Beach morphology and its dynamism from remote sensing for coastal management support / Carlos Cabezas Rabadán (2021)
![]()
Titre : Beach morphology and its dynamism from remote sensing for coastal management support Type de document : Thèse/HDR Auteurs : Carlos Cabezas Rabadán, Auteur ; Josep E. Pardo Pascual, Directeur de thèse ; Miguel Rodilla Alamá, Directeur de thèse Editeur : Valencia : Universitat politécnica de Valencia Année de publication : 2021 Importance : 188 p. Format : 21 x 30 cm Note générale : bibliographie
Thesis dissertation submitted in fulfillment of the requirements for the Degree of Doctor of Philosophy at Universitat Politècnica de ValènciaLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection
[Termes IGN] changement climatique
[Termes IGN] détection de changement
[Termes IGN] érosion côtière
[Termes IGN] géomorphologie
[Termes IGN] image Landsat
[Termes IGN] image Sentinel-MSI
[Termes IGN] modélisation spatio-temporelle
[Termes IGN] plage
[Termes IGN] sédiment
[Termes IGN] surveillance du littoral
[Termes IGN] trait de côteRésumé : (auteur) Beaches are coastal spaces that perform numerous environmental functions. They provide important benefits to society and coastal communities, including the ecological function, the provision of protection for coastal territories, and constitute a basic resource for the tourism industry. Due to climate change and human actions that alter the natural dynamism of the coast, beaches are experiencing increasingly harmful erosive processes that affect their physical integrity and the maintenance of their ecological functions. Beach management is often not adapted to the particularities of the different coastal segments. Decision-making is not based on sufficient information about characteristics, dynamism, and current state of beaches, resulting in short or ineffective solutions. Geomorphological characteristics are essential in the development of beach functions as they condition their physical dimensions and their behavior in response to the action of the sea. Therefore, their detailed and updated characterization is necessary to carry out efficient actions, allowing a more ecosystemic and sustainable coastal management. Remote sensing techniques have a great capacity for acquiring data from the land surface. In particular, Sentinel-2 and Landsat (5, 7, and 8) satellites freely provide medium resolution images with global coverage and high-revisit frequency. The algorithms for extracting the water/land interface recently developed by the Geo-Environmental Cartography and Remote Sensing Group (CGAT – UPV) allow defining the position of the shoreline on these images, constituting potentially useful data to describe beach morphology and dynamics. Universalizing their application requires testing and validation at different coastal types. For this purpose, the extraction process has been adapted for exploitation in tidal environments, and the resulting shorelines have been assessed under different oceanographic conditions offering an accuracy close to 5 m RMSE (Root-Mean-Square Error). From these shorelines, and taking into account the existing information needs for management, it is proposed to derive indicators to characterize the geomorphology of the beaches and to monitor their changes. To this end, the proposed methodologies ensure the efficient management of large volumes of shorelines, being able to characterize the beaches along broad coastal segments and periods. Thus, beach width and sediment grain size are derived as objective and easily understandable indicators of the beach geomorphology. Spatial-temporal modeling of the state and changes of shoreline position and beach width makes it possible to monitor the response to storms and anthropogenic actions, allowing to analyze changes that occur every few days or over decades. The large spatial coverage together with the integration with other cartographic databases allows characterizing the influence of beach geomorphology in the performance of its functions, offering a holistic view of the coast from a regional scale. The methodologies developed in this thesis and the indicators derived from remote sensing provide support and criteria for prioritizing the actions of managers. This contributes to fill the gap between the availability of techniques to obtain remote information and its application in the coastal decision-making process. Note de contenu : 1- General introduction
2- Assessing user’s expectations and perceptions on different beach types and the need for diverse management frameworks
3- Satellite-derived shorelines at an exposed mesotidal beach
4- Characterizing beach changes using satellite-derived shorelines
5- Detecting problematic beach widths for the recreational function from subpixel shoreline
6- Shoreline variability from Sentinel-2: an approach for estimating beach sediment size?
7- Conclusions, management implications and future perspectivesNuméro de notice : 28599 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse étrangère Note de thèse : PhD Thesis : Geomatics : Valencia, Spain : 2021 DOI : 10.4995/Thesis/10251/165076 En ligne : https://doi.org/10.4995/Thesis/10251/165076 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99405
Titre : Benefiting from local rigidity in 3D point cloud processing Type de document : Thèse/HDR Auteurs : Zan Gojcic, Auteur Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2021 Importance : 141 p. Format : 21 x 30 cm Note générale : bibliographie
A thesis submitted to attain the degree of Doctor of Sciences of ETH ZurichLangues : Français (fre) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] apprentissage profond
[Termes IGN] capteur actif
[Termes IGN] champ vectoriel
[Termes IGN] déformation d'image
[Termes IGN] données lidar
[Termes IGN] effondrement de terrain
[Termes IGN] enregistrement de données
[Termes IGN] filtrage du bruit
[Termes IGN] flux
[Termes IGN] image 3D
[Termes IGN] navigation autonome
[Termes IGN] orientation du capteur
[Termes IGN] segmentation
[Termes IGN] semis de points
[Termes IGN] téléphone intelligent
[Termes IGN] traitement de semis de points
[Termes IGN] voxelIndex. décimale : THESE Thèses et HDR Résumé : (auteur) Incorporating 3D understanding and spatial reasoning into (intelligent) algorithms is crucial for solving several tasks in fields such as engineering geodesy, risk assessment, and autonomous driving. Humans are capable of reasoning about 3D spatial relations even from a single 2D image. However, making the priors that we rely on explicit and integrating them into computer programs is very challenging. Operating directly on 3D input data, such as 3D point clouds, alleviates the need to lift 2D data into a 3D representation within the task-specific algorithm and hence reduces the complexity of the problem. The 3D point clouds are not only a better-suited input data representation, but they are also becoming increasingly easier to acquire. Indeed, nowadays, LiDAR sensors are even integrated into consumer devices such as mobile phones. However, these sensors often have a limited field of view, and hence multiple acquisitions are required to cover the whole area of interest. Between these acquisitions, the sensor has to be moved and pointed in a different direction. Moreover, the world that surrounds us is also dynamic and might change as well. Reasoning about the motion of both the sensor and the environment, based on point clouds acquired in two-time steps, is therfore an integral part of point cloud processing. This thesis focuses on incorporating rigidity priors into novel deep learning based approaches for dynamic 3D perception from point cloud data. Specifically, the tasks of point cloud registration, deformation analysis, and scene flow estimation are studied. At first, these tasks are incorporated into a common framework where the main difference is in the level of rigidity assumptions that are imposed on the motion of the scene or
the acquisition sensor. Then, the tasks specific priors are proposed and incorporated into novel deep learning architectures. While the global rigidity can be assumed in point cloud registration, the motion patterns in deformation analysis and scene flow estimation are more complex. Therefore, the global rigidity prior has to be relaxed to local or instancelevel rigidity, respectively. Rigidity priors not only add structure to the aforementioned tasks, which prevents physically implausible estimates and improves the generalization of the algorithms, but in some cases also reduce the supervision requirements. The proposed approaches were quantitatively and qualitatively evaluated on several datasets, and they yield favorable performance compared to the state-of-the-art.Numéro de notice : 28660 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse étrangère Note de thèse : PhD : Sciences : ETH Zurich : 2021 DOI : sans En ligne : https://www.research-collection.ethz.ch/handle/20.500.11850/523368 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99817 Building extraction from Lidar data using statistical methods / Haval Abdul-Jabbar Sadeq in Photogrammetric Engineering & Remote Sensing, PERS, vol 87 n° 1 (January 2021)
![]()
[article]
Titre : Building extraction from Lidar data using statistical methods Type de document : Article/Communication Auteurs : Haval Abdul-Jabbar Sadeq, Auteur Année de publication : 2021 Article en page(s) : pp 33 - 42 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] analyse de données
[Termes IGN] classification orientée objet
[Termes IGN] détection du bâti
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] extraction de traits caractéristiques
[Termes IGN] Ransac (algorithme)
[Termes IGN] segmentation sémantique
[Termes IGN] semis de pointsRésumé : (Auteur) In this article, a straightforward, intuitive method for lidar data classification and building extraction, based on statistical analysis, is presented. The classification of the point cloud into ground and nonground is begun by individually testing each point within the point cloud using the statistical mean height. In this operation, various window sizes are specified, and the mean is obtained at each size. The points that are above the mean are saved and divided by the number of windows to obtain the proportion. Points are considered non-ground if their proportion is higher than the assigned threshold, and otherwise ground. An algorithm for classifying the obtained nonground point cloud into buildings and trees is also illustrated in this article. First the nonground points are labeled, then each label is tested individually. The process begins with segmenting each label. Then comes testing of whether each segment of points can be fitted within a specific plane. The label of the point cloud is considered a building if the number of segments considered as planes is larger than those considered as nonplanes; otherwise it is classified as a tree. Numéro de notice : A2021-055 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.14358/PERS.87.1.33 Date de publication en ligne : 01/01/2021 En ligne : https://doi.org/10.14358/PERS.87.1.33 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=96760
in Photogrammetric Engineering & Remote Sensing, PERS > vol 87 n° 1 (January 2021) . - pp 33 - 42[article]Réservation
Réserver ce documentExemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 105-2021011 SL Revue Centre de documentation Revues en salle Disponible Calcul de la largeur à pleins bords de grands cours d’eau à partir de MNT LiDAR / Nicolas Fermen (2021)
![]()
Titre : Calcul de la largeur à pleins bords de grands cours d’eau à partir de MNT LiDAR Type de document : Mémoire Auteurs : Nicolas Fermen, Auteur Editeur : Le Mans : Ecole Supérieure des Géomètres et Topographes ESGT Année de publication : 2021 Importance : 75 p. Format : 21 x 30 cm Note générale : bibliographie
Mémoire présenté en vue d'obtenir le diplôme d'ingénieur ESGT, spécialité Géomètre et TopographeLangues : Français (fre) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] cartographie automatique
[Termes IGN] Cher (rivière)
[Termes IGN] cours d'eau
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] hydrologie
[Termes IGN] lit majeur
[Termes IGN] modèle numérique de terrain
[Termes IGN] semis de pointsIndex. décimale : ESGT Mémoires d'ingénieurs de l'ESGT Résumé : (auteur) À partir de la première moitié du XIXe siècle, les activités humaines en milieu aquatique (canalisation, barrage, extractions de granulats) n’ont cessé de s’intensifier, engendrant les pressions anthropiques aujourd’hui bien connues. L’étude hydromorphologique du lit des cours d’eau, à travers la géométrie à plein bord (niveau, largeur), permet d’analyser leur réponse érosive et leur adaptation aux divers facteurs perturbant leur dynamique. Il s’agit d’indicateurs essentiels à la compréhension de l’équilibre écologique d’un milieu afin d’assurer le bon état de celui-ci au sens de la Directive Cadre sur l’eau. L’utilisation d’un MNT LiDAR pour l’identification automatique des paramètres à plein bords a permis d’obtenir une cartographie continue des largeurs sur l’ensemble d’une plaine alluviale de grand cours d’eau français. Une approche par profondeur hydraulique reprenant les principes décrits par Faux et al (2009) et De Rosa et al (2019) a été mise en place. L’introduction de nouveaux critères de sélection du niveau à plein bords semble permettre d’appliquer cette méthodologie de manière efficace sur des sites à morphologie complexe. Note de contenu : Introduction
1- Contexte scientifique et sites d’études
2- Méthodes
3- Résultats, discussion et perspectives
ConclusionNuméro de notice : 15287 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Mémoire ingénieur ESGT Organisme de stage : Laboratoire Géomatique et Foncier En ligne : https://dumas.ccsd.cnrs.fr/dumas-03545779/document Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101471 Change detection of land use and land cover, using landsat-8 and sentinel-2A images / Mohammed Abdulmohsen Alhedyan (2021)
PermalinkCombining deep learning and mathematical morphology for historical map segmentation / Yizi Chen (2021)
PermalinkPermalinkConnecting images through time and sources: Introducing low-data, heterogeneous instance retrieval / Dimitri Gominski (2021)
PermalinkConsolidation of crowd-sourced geo-ragged data for parameterized travel recommendations / Ago Luberg (2021)
PermalinkPermalinkContribution des SIG et de la modélisation volumique à la caractérisation géomorphologique et géologique de la région des Doukkala « Meseta côtière, Maroc » / Youness Ahmed Laaziz (2021)
PermalinkContributions to graph-based hierarchical analysis for images and 3D point clouds / Leonardo Gigli (2021)
PermalinkCorrection radiométrique et recalage de nuages de points pour la reconstruction tridimensionnelle d'oeuvres du patrimoine culturel / Nathan Sanchiz (2021)
PermalinkDeep convolutional neural networks for scene understanding and motion planning for self-driving vehicles / Abdelhak Loukkal (2021)
PermalinkPermalinkPermalinkDeep learning for wildfire progression monitoring using SAR and optical satellite image time series / Puzhao Zhang (2021)
PermalinkDescription et recherche d’image généralisables pour l’interconnexion et l’analyse multi-source / Dimitri Gominski (2021)
PermalinkDétection de changement d’occupation du sol à l’aide de données Sentinel en contexte tropical / Lucas Martelet (2021)
PermalinkDétection/reconnaissance d'objets urbains à partir de données 3D multicapteurs prises au niveau du sol, en continu / Younes Zegaoui (2021)
PermalinkDétection et reconstruction 3D d’arbres urbains par segmentation de nuages de points : apport de l’apprentissage profond / Victor Alteirac (2021)
PermalinkPermalinkDéveloppement d’outils d’exploitation des archives photographiques aériennes de l’IGN pour caractériser l’évolution pluridécennale du littoral sur l’île de la Réunion / Adinane Oladjidé Ayichemi (2021)
PermalinkPermalink