Descripteur
Documents disponibles dans cette catégorie (1401)
Ajouter le résultat dans votre panier
Visionner les documents numériques
Affiner la recherche Interroger des sources externes
Etendre la recherche sur niveau(x) vers le bas
Caractérisation de la ville du futur dans des corpus de science-fiction et de fiction climatique / Sami Guembour (2022)
Titre : Caractérisation de la ville du futur dans des corpus de science-fiction et de fiction climatique Type de document : Mémoire Auteurs : Sami Guembour, Auteur ; Catherine Dominguès , Encadrant ; Chuanming Dong , Encadrant Editeur : Paris : Université Paris Cité Année de publication : 2022 Projets : PARVIS / Importance : 53 p. Note générale : bibliographie
Rapport de stage Master 2 informatique, parcours Apprentissage Machine pour la Science des DonnéesLangues : Français (fre) Descripteur : [Vedettes matières IGN] Linguistique
[Termes IGN] apprentissage automatique
[Termes IGN] modèle de langue
[Termes IGN] traitement du langage naturelMots-clés libres : plongement lexical CamemBERT science-fiction embedding Résumé : (auteur) La ville future est souvent décrite dans les romans de science-fiction de fiction climatique de façons innovantes et inventives par les écrivains, et avec la variété des imaginations des auteurs et la multiplicité des romans, la caractérisation de la ville de demain devient compliquée. Le traitement automatique des langues (TAL) est un domaine qui permet de traiter le langage humain automatiquement. Dans ce stage, nous avons eu recours aux techniques et méthodes des sciences des données et du TAL et aux modèles de langue fondés sur les transformers pour classifier les romans de la ville et identifier les différents caractères de la ville du futur et les lieux (en tant que zones socialement reconnues et distinguées) publics et privés et les usages associés. Ce travail a permis de contribuer à la caractérisation de la ville future et les résultats seront valorisés par le projet PARVIS, il a également permis de créer des modèles pour le traitement de la polysémie des mots qui désignent la ville, et sur le plan personnel, il m'a permis d'enrichir mes connaissances en TAL et en science des données et de mieux maîtriser les modèles de langues pour la réalisation des différentes tâches. // The future city is often depicted in climate fiction science fiction novels in innovative and inventive ways by writers, and with the variety of authors’ imaginations and the multiplicity of novels, characterizing the city of tomorrow becomes complicated. Natural language processing (NLP) is a field that allows human language to be processed automatically. In this internship we have used the techniques and methods of data science and NLP and language models based on transformers to classify the novels of the city and identify the different characteristics of the city of the future and the different places (as socially recognized and distinguished areas) public and private and associated uses. This work allowed the characterization of the future city and the results were valued by the PARVIS project, it also made it possible to create models for the treatment of the polysemy of the words which designate the city, and on a personal level it allowed to enrich my knowledge in NLP and data science, and to better master the language models for the realization of the different tasks. Note de contenu : Introduction Générale
1 Contexte du stage
1.1 Présentation de l’organisme d’accueil
1.2 Objectif et étapes du stage
2 Etat de l’art
2.1 Introduction
2.2 Généralités sur le traitement Automatique des Langues
2.3 Domaines d’application
2.4 Différentes étapes du TAL
2.5 Les modèles de langues
2.6 Apprentissage automatique
2.7 Apprentissage profond
2.8 Co-clustering
2.9 Analyse factorielle
2.10 Conclusion
3 Travail réalisé
3.1 Introduction
3.2 Construction du corpus de la ville
3.3 Identification des fonctions associées aux lieux de la ville
3.4 Identification et analyse en sentiments des lieux inventés de la ville
3.5 Conclusion
ConclusionNuméro de notice : 14196 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Mémoire masters divers Organisme de stage : LASTIG (IGN) Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102271 Documents numériques
peut être téléchargé
Rapport de stage de Sami GUEMBOUR - pdf auteurAdobe Acrobat PDF Classification of mediterranean shrub species from UAV point clouds / Juan Pedro Carbonell-Rivera in Remote sensing, vol 14 n° 1 (January-1 2022)
[article]
Titre : Classification of mediterranean shrub species from UAV point clouds Type de document : Article/Communication Auteurs : Juan Pedro Carbonell-Rivera, Auteur ; Jesus Torralba, Auteur ; Javier Estornell, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 199 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques
[Termes IGN] apprentissage automatique
[Termes IGN] arbuste
[Termes IGN] classification par forêts d'arbres décisionnels
[Termes IGN] classification par Perceptron multicouche
[Termes IGN] Espagne
[Termes IGN] Extreme Gradient Machine
[Termes IGN] forêt méditerranéenne
[Termes IGN] image captée par drone
[Termes IGN] incendie de forêt
[Termes IGN] indice de végétation
[Termes IGN] modèle de simulation
[Termes IGN] modèle numérique de terrain
[Termes IGN] parc naturel
[Termes IGN] photogrammétrie aérienne
[Termes IGN] semis de pointsRésumé : (auteur) Modelling fire behaviour in forest fires is based on meteorological, topographical, and vegetation data, including species’ type. To accurately parameterise these models, an inventory of the area of analysis with the maximum spatial and temporal resolution is required. This study investigated the use of UAV-based digital aerial photogrammetry (UAV-DAP) point clouds to classify tree and shrub species in Mediterranean forests, and this information is key for the correct generation of wildfire models. In July 2020, two test sites located in the Natural Park of Sierra Calderona (eastern Spain) were analysed, registering 1036 vegetation individuals as reference data, corresponding to 11 shrub and one tree species. Meanwhile, photogrammetric flights were carried out over the test sites, using a UAV DJI Inspire 2 equipped with a Micasense RedEdge multispectral camera. Geometrical, spectral, and neighbour-based features were obtained from the resulting point cloud generated. Using these features, points belonging to tree and shrub species were classified using several machine learning methods, i.e., Decision Trees, Extra Trees, Gradient Boosting, Random Forest, and MultiLayer Perceptron. The best results were obtained using Gradient Boosting, with a mean cross-validation accuracy of 81.7% and 91.5% for test sites 1 and 2, respectively. Once the best classifier was selected, classified points were clustered based on their geometry and tested with evaluation data, and overall accuracies of 81.9% and 96.4% were obtained for test sites 1 and 2, respectively. Results showed that the use of UAV-DAP allows the classification of Mediterranean tree and shrub species. This technique opens a wide range of possibilities, including the identification of species as a first step for further extraction of structure and fuel variables as input for wildfire behaviour models. Numéro de notice : A2022-057 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article DOI : 10.3390/rs14010199 En ligne : https://doi.org/10.3390/rs14010199 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99462
in Remote sensing > vol 14 n° 1 (January-1 2022) . - n° 199[article]Combining a class-weighted algorithm and machine learning models in landslide susceptibility mapping: A case study of Wanzhou section of the Three Gorges Reservoir, China / Huijuan Zhang in Computers & geosciences, vol 158 (January 2022)
[article]
Titre : Combining a class-weighted algorithm and machine learning models in landslide susceptibility mapping: A case study of Wanzhou section of the Three Gorges Reservoir, China Type de document : Article/Communication Auteurs : Huijuan Zhang, Auteur ; Yingxu Song, Auteur ; Shiluo Xu, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 104966 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Analyse spatiale
[Termes IGN] aléa
[Termes IGN] apprentissage automatique
[Termes IGN] base de données localisées
[Termes IGN] cartographie des risques
[Termes IGN] Chine
[Termes IGN] classification par forêts d'arbres décisionnels
[Termes IGN] effondrement de terrain
[Termes IGN] modèle de simulation
[Termes IGN] régression géographiquement pondérée
[Termes IGN] régression logistique
[Termes IGN] risque naturel
[Termes IGN] Trois Gorges, barrage desRésumé : (auteur) This study aims to investigate the application of a class-weighted algorithm combined with conventional machine learning model (logistic regression (LR)) and ensemble machine learning models (LightGBM and random forest (RF)) to the landslide susceptibility evaluation. Wanzhou section of the Three Gorges Reservoir area, China, frequently suffering numerous landslides, is chosen as an example. The class-weighted algorithm focuses on the class-imbalanced issue of landslide and non-landslide samples, and it can turn the class-imbalanced issue into a cost-sensitive machine learning by setting unequal weights for different classes, which contribute to improving the accuracy of landslide susceptibility evaluation. The landslide inventory database was produced by field investigation and remote sensing images derived from Google Earth. Of the 233 landslides in the inventory, 40% were used for validation, and the remaining 60% were used for training purposes. Twelve environmental parameters (elevation, slope, aspect, curvature, distance to river, NDVI, NDWI, rainfall, seismic intensity, land use, TRI, lithology) were treated as inputs of the models to produce a landslide susceptibility map (LSM). The AUC value, Balanced accuracy, and Geometric mean score were utilized to estimate the quality of models. The result shows that the weighted models (weighted logistic regression (WLR), weighted LightGBM (WLightGBM), weighted random forest (WRF) have higher AUC values, Balanced accuracy, and Geometric mean scores than those of unweighted methods, which demonstrates that the weighted models exhibit better than unweighted models, with the WRF model having the best performance. The landslide susceptibility map of the Wanzhou section displays that the high and very high landslide susceptibility zones are mainly distributed on both sides of the river. The insights from this research will be useful for ameliorating the landslide susceptibility mapping and the prevention and mitigation for the Wanzhou section. Numéro de notice : A2022-029 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article DOI : 10.1016/j.cageo.2021.104966Get rights and content Date de publication en ligne : 27/10/2021 En ligne : https://doi.org/10.1016/j.cageo.2021.104966Get rights and content Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99268
in Computers & geosciences > vol 158 (January 2022) . - n° 104966[article]Construction d’un plugin QGIS de détection d’îlots de chaleur urbains à partir d’images satellitaires de type optique / Houssayn Meriche (2022)
Titre : Construction d’un plugin QGIS de détection d’îlots de chaleur urbains à partir d’images satellitaires de type optique Type de document : Mémoire Auteurs : Houssayn Meriche, Auteur Editeur : Champs-sur-Marne : Ecole nationale des sciences géographiques ENSG Année de publication : 2022 Importance : 51 p. Format : 21 x 30 cm Note générale : Bibliographie
Rapport de projet pluridisciplinaire, cycle ING2Langues : Français (fre) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] carte thématique
[Termes IGN] classification dirigée
[Termes IGN] format GeoTIFF
[Termes IGN] ilot thermique urbain
[Termes IGN] module d'extension
[Termes IGN] Montréal (Québec)
[Termes IGN] Python (langage de programmation)
[Termes IGN] QGIS
[Termes IGN] réseau neuronal convolutif
[Termes IGN] segmentation sémantiqueIndex. décimale : PROJET Mémoires : Rapports de projet - stage des ingénieurs de 2e année Résumé : (Auteur) L’Institut des Sciences de l’Environnement (ISE) est une unité multi départementale de l’Université du Québec à Montréal (UQAM) dans laquelle les sciences de l’environnement ont pour objets les problématiques environnementales découlant des interactions entre processus biologiques, physiques, sociaux et humains. Montréal étant connue dans le monde de la recherche pour sa productivité scientifique dans le domaine de l’Intelligence Artificielle, on retrouve au sein de l’UQAM bon nombre d’étudiants mêlant cette discipline à d’autres champs scientifiques dont l’environnement. C’est ainsi que je suis amené à concevoir un plugin qui, couplé à série d’algorithmes faisant intervenir de l’apprentissage profond, permettrait à une étudiante en maîtrise de Géographie de générer des cartes de prédiction d’îlots de chaleur urbains de la ville de Montréal. Cet ensemble d’algorithmes est réalisé à partir du langage de programmation Python, avec pour support du plugin le logiciel QGIS. Celui-ci est destiné à traiter des images au format exclusif GeoTIFF, et nécessite également des connaissances en fabrication de masque (image binaire constituée de 0 et de 1 renseignant sur la pertinence d’exploitation des pixels de l’image GeoTIFF). Note de contenu : Introduction
1. Contexte du projet
1.1 L’Université du Québec à Montréal (UQAM)
1.2 L’environnement de travail
2. Analyse de l’existant
2.1 Autour des îlots de chaleur en milieu urbain
2.2 L’apprentissage automatique appliqué à la Télédétection
3. Construction du plugin
3.1 Côté Plugin
3.2 Côté Classification
3.3 Résultats et discussions
ConclusionNuméro de notice : 26869 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/IMAGERIE Nature : Mémoire de projet pluridisciplinaire Organisme de stage : Laboratoire de télédétection et de SIG du département de Géographie (Université du Québec à Montréal) Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101696 Documents numériques
peut être téléchargé
Construction d’un Plugin QGIS de détection d’îlots de chaleur urbains à partir d’images satellitaires de type optique - pdf auteurAdobe Acrobat PDF Contribution to object extraction in cartography : A novel deep learning-based solution to recognise, segment and post-process the road transport network as a continuous geospatial element in high-resolution aerial orthoimagery / Calimanut-Ionut Cira (2022)
Titre : Contribution to object extraction in cartography : A novel deep learning-based solution to recognise, segment and post-process the road transport network as a continuous geospatial element in high-resolution aerial orthoimagery Type de document : Thèse/HDR Auteurs : Calimanut-Ionut Cira, Auteur Editeur : Madrid [Espagne] : Universidad politécnica de Madrid Année de publication : 2022 Importance : 227 p. Format : 21 x 30 cm Note générale : bibliographie
Thèse de Doctorat en Topographie, Géodésie et cartographie, Universidad politécnica de MadridLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] analyse d'image orientée objet
[Termes IGN] classification par réseau neuronal convolutif
[Termes IGN] extraction du réseau routier
[Termes IGN] image aérienne
[Termes IGN] orthoimage
[Termes IGN] réseau antagoniste génératif
[Termes IGN] réseau neuronal artificiel
[Termes IGN] route
[Termes IGN] segmentation sémantiqueIndex. décimale : THESE Thèses et HDR Résumé : (auteur) Remote sensing imagery combined with deep learning strategies is often regarded as an ideal solution for interpreting scenes and monitoring infrastructures with remarkable performance levels. Remote sensing experts have been actively using deep neural networks to solve object extraction tasks in high-resolution aerial imagery by means of supervised operations. However, the extraction operation is imperfect, due to the nature of remotely sensed images (noise, obstructions, etc.), the limitations of sensing resolution, or the occlusions often present in the scenes. The road network plays an important part in transportation and, nowadays, one of the main related challenges is keeping the existent cartographic support up to date. This task can be considered very challenging due to the complex nature of the geospatial object (continuous, with irregular geometry, and significant differences in width). We also need to take into account that secondary roads represent the largest part of the road transport network, but due to the absence of clearly defined edges, and the different spectral signatures of the materials used for pavement, monitoring, and mapping them represents a great effort for public administration, and their extraction is often omitted altogether. We believe that recent advancements in machine vision can enable a successful extraction of the road structures from high-resolution, remotely sensed imagery and a greater automation of the road mapping operation. In this PhD thesis, we leverage recent computer vision advances and propose a deep learning-based end-to-end solution, capable of efficiently extracting the surface area of roads at a large scale. The novel approach is based on a disjoint execution of three different image processing operations (recognition, semantic segmentation, and post-processing with conditional generative learning) within a common framework. We focused on improving the state-of-the-art results for each of the mentioned components, before incorporating the resulting models into the proposed solution architecture. For the recognition operation, we proposed two framework candidates based on convolutional neural networks to classify roads in openly available aerial orthoimages divided in tiles of 256×256 pixels, with a spatial resolution of 0.5 m. The frameworks are based on ensemble learning and transfer learning and combine weak classifiers to leverage the strengths of different state-of-the-art models that we heavily modified for computational efficiency. We evaluated their performance on unseen test data and compared the results with those obtained by the state-of-the-art convolutional neural networks trained for the same task, observing improvements in performance metrics of 2-3%. Secondly, we implemented hybrid semantic segmentation models (where the default backbones are replaced by neural network specialised in image segmentation) and trained them with high-resolution remote sensing imagery and their correspondent ground-truth masks. Our models achieved mean increases in performance metrics of 2.7-3.5%, when compared to the original state-of-the-art semantic segmentation architectures trained from scratch for the same task. The best-performing model was integrated on a web platform that handles the evaluation of large areas, the association of the semantic predictions with geographical coordinates, the conversion of the tiles’ format, and the generation of GeoTIFF results (compatible with geospatial databases). Thirdly, the road surface area extraction task is generally carried out via semantic segmentation over remotely sensed imagery—however, this supervised learning task can be considered very costly because it requires remote sensing images labelled at pixel level and the results are not always satisfactory (presence of discontinuities, overlooked connection points, or isolated road segments). We consider that unsupervised learning (not requiring labelled data) can be employed for post-processing the geometries of geospatial objects extracted via semantic segmentation. For this reason, we also approached the post-processing of the road surface areas obtained with the best performing segmentation model to improve the initial segmentation predictions. In this line, we proposed two post-processing operations based on conditional generative learning for deep inpainting and image-to-image translation operations and trained the networks to learn the distribution of the road network present in official cartography, using a novel dataset covering representative areas of Spain. The first proposed conditional Generative Adversarial Network (cGAN) model was trained for deep inpainting operation and obtained improvements in performance metrics of maximum 1.3%. The second cGAN model was trained for image-to-image translation, is based on a popular model heavily modified for computational efficiency (a 92.4% decrease in the number of parameters in the generator network and a 61.3% decrease in the discriminator network), and achieved a maximum increase of 11.6% in performance metrics. We also conducted a qualitative comparison to visually assess the effectiveness of the generative operations and observed great improvements with respect to the initial semantic segmentation predictions. Lastly, we proposed an end-to-end processing strategy that combines image classification, semantic segmentation, and post-processing operations to extract containing road surface area extraction from high-resolution aerial orthophotography. The training of the model components was carried out on a large-scale dataset containing more than 537,500 tiles, covering approximately 20,800 km2 of the Spanish territory, manually tagged at pixel level. The consecutive execution of the resulting deep learning models delivered higher quality results when compared to state-of-the-art implementations trained for the same task. The versatility and flexibility of the solution given by the disjointed execution of the three separate sub-operations proved its effectiveness and economic efficiency and enables the integration of a web application that alleviates the manipulation of geospatial data, while allowing for an easy integration of future models and algorithms. Resuming, applying the proposed models resulted from this PhD thesis translates to operations aimed to check if the latest existing aerial orthoimages contains the studied continuous geospatial element, to obtain an approximation of its surface area using supervised learning and to improve the initial segmentation results with post-processing methods based on conditional generative learning. The results obtained with the proposed end-to-end-solution presented in this PhD thesis improve the state-of-the-art in the field of road extraction with deep learning techniques and prove the appropriateness of applying the proposed extraction workflow for a more robust and more efficient extraction operation of the road transport network. We strongly believe that the processing strategy can be applied to enhance other similar extraction tasks of continuous geospatial elements (such as the mapping of riverbeds, or railroads), or serve as a base for developing additional extraction workflows of geospatial objects from remote sensing images. Note de contenu : 1- Introduction
2- Methodology
3- Theoretical framework
4- Litterature review
5- Road recognition: A framework based on nestion of convolutional neuronal networks and transfer learning to regognise road elements
6- Road segmentation: An approach based on hybrid semantic segmentation models to extract the surface area of rod elements from aerial orthoimagery
7- Post-processing of semantic segmentation predictions I: A conditional generative adversial network to improve the extraction of road surface areas via deep inpainting operations
8- Post-processing of semantic segmentation predictions II: A lightweight conditional generative adversial network to improve the extraction of road surface areas via image-to-image translation
9- An end-to-end road extraction solution based on regonition, segmentation, and post-processing operations for a large-scale mapping of the road transport network from aerial orthophotography
10- ConclusionsNuméro de notice : 24069 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse étrangère Note de thèse : Thèse de Doctorat : Topographie, Géodésie et cartographie : Universidad politécnica de Madrid : 2022 DOI : 10.20868/UPM.thesis.70152 En ligne : https://doi.org/10.20868/UPM.thesis.70152 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102113 PermalinkCultivating historical heritage area vitality using urban morphology approach based on big data and machine learning / Jiayu Wu in Computers, Environment and Urban Systems, vol 91 (January 2022)PermalinkDeep image translation with an affinity-based change prior for unsupervised multimodal change detection / Luigi Tommaso Luppino in IEEE Transactions on geoscience and remote sensing, vol 60 n° 1 (January 2022)PermalinkPermalinkDeep learning based 2D and 3D object detection and tracking on monocular video in the context of autonomous vehicles / Zhujun Xu (2022)PermalinkPermalinkPermalinkPermalinkPermalinkPermalinkPermalinkPermalinkDetection of windthrown tree stems on UAV-orthomosaics using U-Net convolutional networks / Stefan Reder in Remote sensing, vol 14 n° 1 (January-1 2022)PermalinkDétection des prairies de fauche et estimation des périodes de fauche par télédétection / Emma Seneschal (2022)PermalinkDeveloping the potential of airborne lidar systems for the sustainable management of forests / Karun Dayal (2022)PermalinkDevelopment of object detectors for satellite images by deep learning / Alissa Kouraeva (2022)PermalinkPermalinkEstimating aboveground biomass in dense Hyrcanian forests by the use of Sentinel-2 data / Fardin Moradi in Forests, vol 13 n° 1 (January 2022)PermalinkÉvaluation de la qualité des données géographiques d'OpenStreetMap à l'aide des méthodes d'apprentissage automatique : cas de la République de Djibouti / Ibrahim Maidaneh Abdi (2022)PermalinkExploring data fusion for multi-object detection for intelligent transportation systems using deep learning / Amira Mimouna (2022)Permalink