Descripteur
Termes IGN > informatique > intelligence artificielle > apprentissage automatique > apprentissage profond
apprentissage profond |
Documents disponibles dans cette catégorie (748)
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
Titre : Multi-layer modeling of dense vegetation from aerial LiDAR scans Type de document : Article/Communication Auteurs : Ekaterina Kalinicheva , Auteur ; Loïc Landrieu , Auteur ; Clément Mallet , Auteur ; Nesrine Chehata , Auteur Editeur : Computer vision foundation CVF Année de publication : 2022 Projets : 1-Pas de projet / Conférence : EarthVision 2022, Large Scale Computer Vision for Remote Sensing Imagery, workshop joint to CVPR 2022 19/06/2022 24/06/2022 New Orleans Louisiane - Etats-Unis OA Proceedings Importance : pp 1341 - 1350 Format : 21 x 30 cm Note générale : bibliographie Langues : Français (fre) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] apprentissage profond
[Termes IGN] canopée
[Termes IGN] carte d'occupation du sol
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] étage de végétation
[Termes IGN] foresterie
[Termes IGN] maillage
[Termes IGN] parcelle forestière
[Termes IGN] reconstruction d'objet
[Termes IGN] segmentation d'image
[Termes IGN] semis de pointsRésumé : (auteur) The analysis of the multi-layer structure of wild forests is an important challenge of automated large-scale forestry. While modern aerial LiDARs offer geometric information across all vegetation layers, most datasets and methods focus only on the segmentation and reconstruction of the top of canopy. We release WildForest3D, which consists of 29 study plots and over 2000 individual trees across 47 000m2 with dense 3D annotation, along with occupancy and height maps for 3 vegetation layers: ground vegetation, understory, and overstory. We propose a 3D deep net- work architecture predicting for the first time both 3D point- wise labels and high-resolution layer occupancy rasters simultaneously. This allows us to produce a precise estimation of the thickness of each vegetation layer as well as the corresponding watertight meshes, therefore meeting most forestry purposes. Both the dataset and the model are released in open access: https://github.com/ ekalinicheva/multi_layer_vegetation. Numéro de notice : C2022-007 Affiliation des auteurs : UGE-LASTIG+Ext (2020- ) Autre URL associée : vers CVF Thématique : FORET/IMAGERIE Nature : Communication nature-HAL : ComAvecCL&ActesPubliésIntl DOI : 10.1109/CVPRW56347.2022.00140 Date de publication en ligne : 25/04/2022 En ligne : https://arxiv.org/abs/2204.11620 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100509 Optimization of deep neural networks: A functional perspective with applications in image classification / Simon Roburin (2022)
Titre : Optimization of deep neural networks: A functional perspective with applications in image classification Type de document : Thèse/HDR Auteurs : Simon Roburin, Auteur ; Mathieu Aubry, Directeur de thèse Editeur : Champs-sur-Marne : Ecole des Ponts ParisTech Année de publication : 2022 Importance : 141 p. Format : 21 x 30 cm Note générale : Bibliographie
Thèse de Doctorat de l'Ecole des Ponts ParisTech, spécialité Mathématiques AppliquéesLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] analyse de groupement
[Termes IGN] apprentissage profond
[Termes IGN] classification par nuées dynamiques
[Termes IGN] mathématiques appliquées
[Termes IGN] optimisation (mathématiques)
[Termes IGN] vision par ordinateurIndex. décimale : THESE Thèses et HDR Résumé : (auteur) Despite numerous successes in a wide range of industrial and scientific applications, the learning process of deep neural networks is poorly understood. Loosely speaking, learning aims at finding the network parameters that not only minimize the network errors on a set of training examples but also yield correct predictions on unseen data. Under the prism of optimization, it boils down to minimizing a high dimensional non-convex function. Generalization can generally be expected when one has access to very large datasets and assumes that both training examples and unseen data are sampled from identically independently distributed random variables. The goal of this thesis is to develop analytical tools to better understand neural network optimization and to improve the design of training algorithms in the context of image classification. Note de contenu : 1- Introduction
2- Literature review
3- Impact of Normalization Layers on Optimization
4- Avoid learning spurious correlations
5- ConclusionNuméro de notice : 24098 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Thèse française Note de thèse : Thèse de Doctorat : Mathématiques Appliquées : Ponts ParisTech : 2022 Organisme de stage : LIGM-IMAGINE En ligne : https://hal.science/tel-03968114v1 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102573 Photogrammetric point clouds: quality assessment, filtering, and change detection / Zhenchao Zhang (2022)
Titre : Photogrammetric point clouds: quality assessment, filtering, and change detection Type de document : Thèse/HDR Auteurs : Zhenchao Zhang, Auteur ; M. George Vosselman, Auteur ; Markus Gerke, Auteur ; Michael Ying Yang, Auteur Editeur : Enschede [Pays-Bas] : International Institute for Geo-Information Science and Earth Observation ITC Année de publication : 2022 Note générale : bibliographie
NB : EMBARGO SUR LE TEXTE JUSQU'AU 1ER JUILLET 2022Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques
[Termes IGN] appariement dense
[Termes IGN] détection de changement
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] qualité des données
[Termes IGN] réseau neuronal convolutif
[Termes IGN] segmentation sémantique
[Termes IGN] semis de pointsRésumé : (auteur) 3D change detection draws more and more attention in recent years due to the increasing availability of 3D data. It can be used in the fields of land use / land cover (LULC) change detection, 3D geographic information updating, terrain deformation analysis, urban construction monitoring et al. Our motivation to study 3D change detection is mainly related to the practical need to update the outdated point clouds captured by Airborne Laser Scanning (ALS) with new point clouds obtained by dense image matching (DIM).
The thesis has three main parts. The first part, chapter 1, explains the motivation, providing a review of current ALS and airborne photogrammetry techniques. It also presents the research objectives and questions. The second part including chapter 2 and chapter 3 evaluates the quality of photogrammetric products and investigates their potential for change detection. The third part including chapter 4 and chapter 5 proposes two methods for change detection that meet different requirements.
To investigate the potential of using point clouds derived by dense matching for change detection, we propose a framework for evaluating the quality of 3D point clouds and DSMs generated by dense image matching. Our evaluation framework based on a large number of square patches reveals the distribution of dense matching errors in the whole photogrammetric block. Robust quality measures are proposed to indicate the DIM accuracy and precision quantitatively. The overall mean offset to the reference is 0.1 Ground Sample Distance (GSD); the maximum mean deviation reaches 1.0 GSD. We also find that the distribution of dense matching errors is homogenous in the whole block and close to a normal distribution based on many patch-based samples. However, in some locations, especially along narrow alleys, the mean deviations may get worse. In addition, the profiles of ALS points and DIM points reveal that the DIM profile fluctuates around the ALS profile. We find that the accuracy of DIM point cloud improves and that the noise level decreases on smooth ground areas when oblique images are used in dense matching together with nadir images.
Then we evaluate whether the standard LiDAR filters are effective to filter dense matching points in order to derive accurate DTMs. Filtering results on a city block show that LiDAR filters perform well on the grassland, along bushes and around individual trees if the point cloud is sufficiently precise. When a ranking filter is used on the point clouds before filtering, the filtering will identify fewer but more reliable ground points. However, some small objects on the terrain will be filtered out. Since we aim at obtaining accurate DTMs, the ranking filter shows its value in identifying reliable ground points. Based on the previous findings in DIM quality, we propose a method to detect building changes between ALS and photogrammetric data. Firstly, the ALS points and DIM points are split out and concatenated with the orthoimages. The multimodal data are normalized to feed into a pseudo-Siamese Neural network for change detection. Then, the changed objects are delineated through per-pixel classification and artefact removal. The change detection module based on a pseudo-Siamese CNN can quickly localize the changes and generate coarse change maps. The next module can be used in precise mapping of change boundaries. Experimental results show that the proposed pseudo-Siamese Neural network can cope with the DIM errors and output plausible change detection results. Although the point cloud quality from dense matching is not as fine as laser scanning points, the spectral and textural information provided by the orthoimages serve as a supplement.
Considering that the tasks of semantic segmentation and change detection are correlated, we propose SiamPointNet++ model to combine the two tasks in one framework. The method outputs a pointwise joint label for each ALS point. If an ALS point is unchanged, it is assigned a semantic label; If an ALS point is changed, it is assigned a change label. The sematic and change information are included in the joint labels with minimum information redundancy. The combined Siamese network learns both intra-epoch and inter-epoch features. Intra-epoch features are extracted at multiple scales to embed the local and global information. Inter-epoch features are extracted by Conjugated Ball Sampling (CBS) and concatenated to make change inference. Experiments on the Rotterdam data set indicate that the network is effective in learning multi-task features. It is invariant to the permutation and noise of inputs and robust to the data difference between ALS and DIM data. Compared with a sophisticated object-based method and supervised change detection, this method requires much less hyper-parameters and human intervention but achieves superior performance.
As a conclusion, the thesis evaluates the quality of dense matching points and investigates its potential of updating outdated ALS points. The two change detection methods developed for different applications show their potential in the automation of topographic change detection and point cloud updating. Future work may focus on improving the generalizability and interpretability of the proposed models.Numéro de notice : 20403 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Thèse étrangère Note de thèse : PhD thesis : Geo-Information Science and Earth Observation : Enschede, university of Twente : 2022 DOI : 10.3990/1.9789036552653 Date de publication en ligne : 14/01/2022 En ligne : https://research.utwente.nl/en/publications/photogrammetric-point-clouds-quality [...] Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100963 Représentation et combinaison de l'information géographique pour l'apprentissage profond / Azelle Courtial (2022)
Titre : Représentation et combinaison de l'information géographique pour l'apprentissage profond Type de document : Article/Communication Auteurs : Azelle Courtial , Auteur ; Guillaume Touya , Auteur Editeur : [s.l.] : [s.n.] Année de publication : 2022 Conférence : EGC 2022, 7e atelier GAST, Gestion et Analyse de données Spatiales et Temporelles 25/01/2022 25/01/2022 Blois France OA Proceedings Importance : pp 54 - 65 Note générale : bibliographie Langues : Français (fre) Descripteur : [Termes IGN] apprentissage profond
[Termes IGN] données d'entrainement (apprentissage automatique)
[Termes IGN] données vectorielles
[Termes IGN] généralisation automatique de données
[Termes IGN] représentation des données
[Termes IGN] tenseur
[Vedettes matières IGN] GénéralisationRésumé : (auteur) L’apprentissage profond permet maintenant de générer des cartes transformées à partir d’images d’autres cartes. Mais contrairement aux méthodes traditionnelles de prédiction de carte qui reposent sur des couches de données vectorielles stockées dans des bases de données géographiques, l’image ne transmet qu’une vue limitée des informations contenues dans la version vectorielle des données. Dans cet article, nous nous intéressons à la représentation de l’information géographique sous forme de tenseurs pour améliorer la génération de cartes par apprentissage profond. Nous proposons d’abord une stratégie alternative pour la création des données d’apprentissage : un ensemble de masques où chacun décrit les formes et positions d’un type d’objet géographique sur une même portion de carte (bâtiments, routes, ...). Nous étudions ensuite comment combiner de l’information géographique additionnelle dans les mécanismes d’apprentissage pour améliorer l’abstraction des cartes générées. Numéro de notice : C2022-054 Affiliation des auteurs : UGE-LASTIG (2020- ) Thématique : GEOMATIQUE/INFORMATIQUE Nature : Communication nature-HAL : ComAvecCL&ActesPubliésIntl DOI : sans En ligne : https://hal.science/hal-03719234v1 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=103160 Documents numériques
peut être téléchargé
Représentation et combinaison de l'IG ... - pdf éditeurAdobe Acrobat PDF Representing vector geographic information as a tensor for deep learning based map generalisation / Azelle Courtial (2022)
Titre : Representing vector geographic information as a tensor for deep learning based map generalisation Type de document : Article/Communication Auteurs : Azelle Courtial , Auteur ; Guillaume Touya , Auteur ; Xiang Zhang, Auteur Editeur : AGILE Alliance Année de publication : 2022 Projets : 1-Pas de projet / Conférence : AGILE 2022, 25th international AGILE Conference on Geographic Information Science, Artificial intelligence in the service of geospatial technologies 14/06/2022 17/06/2022 Vilnius Lithuanie OA Proceedings Importance : 8 p. Format : 21 x 30 cm Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] alignement des données
[Termes IGN] apprentissage profond
[Termes IGN] architecture de réseau
[Termes IGN] bati
[Termes IGN] carte topographique
[Termes IGN] couche
[Termes IGN] données d'entrainement (apprentissage automatique)
[Termes IGN] données vectorielles
[Termes IGN] information sémantique
[Termes IGN] milieu urbain
[Termes IGN] route
[Termes IGN] tenseur
[Vedettes matières IGN] GénéralisationRésumé : (auteur) Recently, many researchers tried to generate (generalised) maps using deep learning, and most of the proposed methods deal with deep neural network architecture choices. Deep learning learns to reproduce examples, so we think that improving the training examples, and especially the representation of the initial geographic information, is the key issue for this problem. Our article extracts some representation issues from a literature review and proposes different ways to represent vector geographic information as a tensor. We propose two kinds of contributions: 1) the representation of information by layers; 2) the representation of additional information. Then, we demonstrate the interest of some of our propositions with experiments that show a visual improvement for the generation of generalised topographic maps in urban areas. Numéro de notice : C2022-024 Affiliation des auteurs : UGE-LASTIG+Ext (2020- ) Autre URL associée : vers HAL Thématique : GEOMATIQUE Nature : Communication nature-HAL : ComAvecCL&ActesPubliésIntl DOI : 10.5194/agile-giss-3-32-2022 En ligne : https://doi.org/10.5194/agile-giss-3-32-2022 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100921 PermalinkPermalinkSelf-attention and generative adversarial networks for algae monitoring / Nhut Hai Huynh in European journal of remote sensing, vol 55 n° 1 (2022)PermalinkSemantic segmentation of high-resolution remote sensing images based on a class feature attention mechanism fused with Deeplabv3+ / Zhimin Wang in Computers & geosciences, vol 158 (January 2022)PermalinkStudying informativeness of satellite image texture for sea ice state retrieval using deep learning methods / Clément Fougerouse (2022)PermalinkTowards expressive graph neural networks : Theory, algorithms, and applications / Georgios Dasoulas (2022)PermalinkTowards synthetic sensing for smart cities : a machine/deep learning-based approach / Faraz Malik Awan (2022)PermalinkUrban infrastructure audit: an effective protocol to digitize signalized intersections by mining street view images / Xiao Li in Cartography and Geographic Information Science, vol 49 n° 1 (January 2022)PermalinkEfficient occluded road extraction from high-resolution remote sensing imagery / Dejun Feng in Remote sensing, vol 13 n° 24 (December-2 2021)PermalinkAutomatic extraction of indoor spatial information from floor plan image: A patch-based deep learning methodology application on large-scale complex buildings / Hyunjung Kim in ISPRS International journal of geo-information, vol 10 n° 12 (December 2021)PermalinkBuilding detection with convolutional networks trained with transfer learning / Simon Šanca in Geodetski vestnik, vol 65 n° 4 (December 2021 - February 2022)PermalinkDeep learning for toponym resolution: Geocoding based on pairs of toponyms / Jacques Fize in ISPRS International journal of geo-information, vol 10 n° 12 (December 2021)PermalinkDiResNet: Direction-aware residual network for road extraction in VHR remote sensing images / Lei Ding in IEEE Transactions on geoscience and remote sensing, vol 59 n° 12 (December 2021)PermalinkA hierarchical deep neural network with iterative features for semantic labeling of airborne LiDAR point clouds / Yetao Yang in Computers & geosciences, vol 157 (December 2021)PermalinkImproving soil moisture retrieval from GNSS-interferometric reflectometry: parameters optimization and data fusion via neural network / Yajie Shi in International Journal of Remote Sensing IJRS, vol 42 n° 23 (1-10 December 2021)PermalinkLithological mapping based on fully convolutional network and multi-source geological data / Ziye Wang in Remote sensing, vol 13 n° 23 (December-1 2021)PermalinkUnderstanding and predicting the spatio-temporal spread of COVID-19 via integrating diffusive graph embedding and compartmental models / Tong Zhang in Transactions in GIS, vol 25 n° 6 (December 2021)PermalinkCrop rotation modeling for deep learning-based parcel classification from satellite time series / Félix Quinton in Remote sensing, vol 13 n° 22 (November-2 2021)PermalinkA comparison of a gradient boosting decision tree, random forests, and artificial neural networks to model urban land use changes: the case of the Seoul metropolitan area / Myung-Jin Jun in International journal of geographical information science IJGIS, vol 35 n° 11 (November 2021)PermalinkDownscaling MODIS spectral bands using deep learning / Rohit Mukherjee in GIScience and remote sensing, vol 58 n° 8 (2021)Permalink