Descripteur
Termes IGN > informatique > intelligence artificielle > apprentissage automatique > apprentissage dirigé
apprentissage dirigéSynonyme(s)apprentissage superviséVoir aussi |
Documents disponibles dans cette catégorie (108)
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
An object-based approach for mapping forest structural types based on low-density LiDAR and multispectral imagery / Luis Angel Ruiz in Geocarto international, vol 33 n° 5 (May 2018)
[article]
Titre : An object-based approach for mapping forest structural types based on low-density LiDAR and multispectral imagery Type de document : Article/Communication Auteurs : Luis Angel Ruiz, Auteur ; Jorge Abel Recio, Auteur ; Pablo Crespo-Peremarch, Auteur ; Marta Sapena Moll, Auteur Année de publication : 2018 Article en page(s) : pp 443 - 457 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques
[Termes IGN] arbre de décision
[Termes IGN] biomasse (combustible)
[Termes IGN] carte forestière
[Termes IGN] classification barycentrique
[Termes IGN] classification orientée objet
[Termes IGN] classification par forêts d'arbres décisionnels
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] forêt méditerranéenne
[Termes IGN] image Sentinel-MSI
[Termes IGN] image Worldview
[Termes IGN] modèle de simulation
[Termes IGN] structure d'un peuplement forestierRésumé : (Auteur) Mapping forest structure variables provides important information for the estimation of forest biomass, carbon stocks, pasture suitability or for wildfire risk prevention and control. The optimization of the prediction models of these variables requires an adequate stratification of the forest landscape in order to create specific models for each structural type or strata. This paper aims to propose and validate the use of an object-oriented classification methodology based on low-density LiDAR data (0.5 m−2) available at national level, WorldView-2 and Sentinel-2 multispectral imagery to categorize Mediterranean forests in generic structural types. After preprocessing the data sets, the area was segmented using a multiresolution algorithm, features describing 3D vertical structure were extracted from LiDAR data and spectral and texture features from satellite images. Objects were classified after feature selection in the following structural classes: grasslands, shrubs, forest (without shrubs), mixed forest (trees and shrubs) and dense young forest. Four classification algorithms (C4.5 decision trees, random forest, k-nearest neighbour and support vector machine) were evaluated using cross-validation techniques. The results show that the integration of low-density LiDAR and multispectral imagery provide a set of complementary features that improve the results (90.75% overall accuracy), and the object-oriented classification techniques are efficient for stratification of Mediterranean forest areas in structural- and fuel-related categories. Further work will be focused on the creation and validation of a different prediction model adapted to the various strata. Numéro de notice : A2018-140 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/10106049.2016.1265595 Date de publication en ligne : 28/11/2016 En ligne : https://doi.org/10.1080/10106049.2016.1265595 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=89690
in Geocarto international > vol 33 n° 5 (May 2018) . - pp 443 - 457[article]Large-scale supervised learning for 3D Point cloud labeling : Semantic3d.Net / Timo Hackel in Photogrammetric Engineering & Remote Sensing, PERS, vol 84 n° 5 (mai 2018)
[article]
Titre : Large-scale supervised learning for 3D Point cloud labeling : Semantic3d.Net Type de document : Article/Communication Auteurs : Timo Hackel, Auteur ; Jan Dirk Wegner, Auteur ; Nikolay Savinov, Auteur ; Lubor Ladicky, Auteur ; Konrad Schindler, Auteur ; Marc Pollefeys, Auteur Année de publication : 2018 Article en page(s) : pp 297 - 308 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] apprentissage dirigé
[Termes IGN] apprentissage profond
[Termes IGN] classification
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] état de l'art
[Termes IGN] réseau neuronal convolutif
[Termes IGN] scène urbaine
[Termes IGN] segmentation sémantique
[Termes IGN] semis de pointsRésumé : (Auteur) In this paper, we review current state-of-the-art in 3D point cloud classification, present a new 3D point cloud classification benchmark data set of single scans with over four billion manually labeled points, and discuss first available results on the benchmark. Much of the stunning recent progress in 2D image interpretation can be attributed to the availability of large amounts of training data, which have enabled the (supervised) learning of deep neural networks. With the data set presented in this paper, we aim to boost the performance of CNNs also for 3D point cloud labeling. Our hope is that this will lead to a breakthrough of deep learning also for 3D (geo-) data. The semantic3D.net data set consists of dense point clouds acquired with static terrestrial laser scanners. It contains eight semantic classes and covers a wide range of urban outdoor scenes, including churches, streets, railroad tracks, squares, villages, soccer fields, and castles. We describe our labeling interface and show that, compared to those already available to the research community, our data set provides denser and more complete point clouds, with a much higher overall number of labeled points. We further provide descriptions of baseline methods and of the first independent submissions, which are indeed based on CNNs, and already show remarkable improvements over prior art. We hope that semantic3D.net will pave the way for deep learning in 3D point cloud analysis, and for 3D representation learning in general. Numéro de notice : A2018-162 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.14358/PERS.84.5.297 Date de publication en ligne : 01/05/2018 En ligne : https://doi.org/10.14358/PERS.84.5.297 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=89795
in Photogrammetric Engineering & Remote Sensing, PERS > vol 84 n° 5 (mai 2018) . - pp 297 - 308[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 105-2018051 RAB Revue Centre de documentation En réserve L003 Disponible Local curvature entropy-based 3D terrain representation using a comprehensive Quadtree / Giyu Chen in ISPRS Journal of photogrammetry and remote sensing, vol 139 (May 2018)
[article]
Titre : Local curvature entropy-based 3D terrain representation using a comprehensive Quadtree Type de document : Article/Communication Auteurs : Giyu Chen, Auteur ; Gang Liu, Auteur ; Xiaogang Ma, Auteur ; Gregoire Mariethoz, Auteur ; Zhenwen He, Auteur ; et al., Auteur Année de publication : 2018 Article en page(s) : pp 30 - 45 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] arbre quadratique
[Termes IGN] entropie
[Termes IGN] modèle numérique de terrain
[Termes IGN] niveau de détail
[Termes IGN] visualisation 3DRésumé : (Auteur) Large scale 3D digital terrain modeling is a crucial part of many real-time applications in geoinformatics. In recent years, the improved speed and precision in spatial data collection make the original terrain data more complex and bigger, which poses challenges for data management, visualization and analysis. In this work, we presented an effective and comprehensive 3D terrain representation based on local curvature entropy and a dynamic Quadtree. The Level-of-detail (LOD) models of significant terrain features were employed to generate hierarchical terrain surfaces. In order to reduce the radical changes of grid density between adjacent LODs, local entropy of terrain curvature was regarded as a measure of subdividing terrain grid cells. Then, an efficient approach was presented to eliminate the cracks among the different LODs by directly updating the Quadtree due to an edge-based structure proposed in this work. Furthermore, we utilized a threshold of local entropy stored in each parent node of this Quadtree to flexibly control the depth of the Quadtree and dynamically schedule large-scale LOD terrain. Several experiments were implemented to test the performance of the proposed method. The results demonstrate that our method can be applied to construct LOD 3D terrain models with good performance in terms of computational cost and the maintenance of terrain features. Our method has already been deployed in a geographic information system (GIS) for practical uses, and it is able to support the real-time dynamic scheduling of large scale terrain models more easily and efficiently. Numéro de notice : A2018-110 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2018.03.001 En ligne : https://doi.org/10.1016/j.isprsjprs.2018.03.001 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=89540
in ISPRS Journal of photogrammetry and remote sensing > vol 139 (May 2018) . - pp 30 - 45[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 081-2018051 RAB Revue Centre de documentation En réserve L003 Disponible Désambiguïsation des entités spatiales par apprentissage actif / Amal Chihaoui in Revue internationale de géomatique, vol 28 n° 2 (avril - juin 2018)
[article]
Titre : Désambiguïsation des entités spatiales par apprentissage actif Type de document : Article/Communication Auteurs : Amal Chihaoui, Auteur ; Asma Bouhafs, Auteur ; Mathieu Roche, Auteur ; Maguelonne Teisseire, Auteur Année de publication : 2018 Article en page(s) : pp 163 - 189 Note générale : Bibliographie Langues : Français (fre) Descripteur : [Vedettes matières IGN] Intelligence artificielle
[Termes IGN] apprentissage dirigé
[Termes IGN] corpus
[Termes IGN] échantillonnage
[Termes IGN] extraction automatique
[Termes IGN] incertitude d'attribut
[Termes IGN] toponyme
[Termes IGN] traitement du langage naturelRésumé : (Auteur) L’extraction de connaissances spatiales à partir de documents textuels peut être une tâche difficile du fait de l’ambiguïté propre au langage naturel. L'indisponibilité de gros volumes de données étiquetées rend difficile la mise-en-œuvre d’un processus de découverte automatique. Dans ce contexte, nous abordons le problème de la désambiguïsation des entités spatiales, entre " localisation" et "organisation" par apprentissage actif. D’abord, nous introduisons une méthode de résolution des toponymes basée sur une analyse lexicale et contextuelle. Ensuite, nous proposons une amélioration en intégrant un modèle d’apprentissage actif. Celui-ci permet de sélectionner automatiquement les données non étiquetées les plus informatives pour la notation humaine. Les expérimentations sont réalisées sur un corpus de "SemEval-2007" en anglais et soulignent l’amélioration du modèle d’apprentissage initial avec un étiquetage réduit. Numéro de notice : A2018-254 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/TOPONYMIE Nature : Article DOI : 10.3166/rig.2018.00053 Date de publication en ligne : 03/08/2018 En ligne : https://doi.org/10.3166/rig.2018.00053 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=90315
in Revue internationale de géomatique > vol 28 n° 2 (avril - juin 2018) . - pp 163 - 189[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 047-2018021 SL Revue Centre de documentation Revues en salle Disponible A spatio-temporal index for aerial full waveform laser scanning data / Debra F. Laefer in ISPRS Journal of photogrammetry and remote sensing, vol 138 (April 2018)
[article]
Titre : A spatio-temporal index for aerial full waveform laser scanning data Type de document : Article/Communication Auteurs : Debra F. Laefer, Auteur ; Anh-Vu Vo, Auteur ; Michela Bertolotto, Auteur Année de publication : 2018 Article en page(s) : pp 232 - 251 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] arbre-R
[Termes IGN] base de données localisées
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] forme d'onde pleine
[Termes IGN] index spatiotemporel
[Termes IGN] indexation spatiale
[Termes IGN] octreeRésumé : (Auteur) Aerial laser scanning is increasingly available in the full waveform version of the raw signal, which can provide greater insight into and control over the data and, thus, richer information about the scanned scenes. However, when compared to conventional discrete point storage, preserving raw waveforms leads to vastly larger and more complex data volumes. To begin addressing these challenges, this paper introduces a novel bi-level approach for storing and indexing full waveform (FWF) laser scanning data in a relational database environment, while considering both the spatial and the temporal dimensions of that data. In the storage scheme's upper level, the full waveform datasets are partitioned into spatial and temporal coherent groups that are indexed by a two-dimensional R∗-tree. To further accelerate intra-block data retrieval, at the lower level a three-dimensional local octree is created for each pulse block. The local octrees are implemented in-memory and can be efficiently written to a database for reuse. The indexing solution enables scalable and efficient three-dimensional (3D) spatial and spatio-temporal queries on the actual pulse data - functionalities not available in other systems. The proposed FWF laser scanning data solution is capable of managing multiple FWF datasets derived from large flight missions. The flight structure is embedded into the data storage model and can be used for querying predicates. Such functionality is important to FWF data exploration since aircraft locations and orientations are frequently required for FWF data analyses. Empirical tests on real datasets of up to 1 billion pulses from Dublin, Ireland prove the almost perfect scalability of the system. The use of the local 3D octree in the indexing structure accelerated pulse clipping by 1.2–3.5 times for non-axis-aligned (NAA) polyhedron shaped clipping windows, while axis-aligned (AA) polyhedron clipping was better served using only the top indexing layer. The distinct behaviours of the hybrid indexing for AA and NAA clipping windows are attributable to the different proportion of the local-index-related overheads with respect to the total querying costs. When temporal constraints were added, generally the number of costly spatial checks were reduced, thereby shortening the querying times. Numéro de notice : A2018-125 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2018.01.012 En ligne : https://doi.org/10.1016/j.isprsjprs.2018.01.012 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=89587
in ISPRS Journal of photogrammetry and remote sensing > vol 138 (April 2018) . - pp 232 - 251[article]Exemplaires(3)
Code-barres Cote Support Localisation Section Disponibilité 081-2018041 RAB Revue Centre de documentation En réserve L003 Disponible 081-2018043 DEP-EXM Revue LASTIG Dépôt en unité Exclu du prêt 081-2018042 DEP-EAF Revue Nancy Dépôt en unité Exclu du prêt Analyse de l'incertitude et de la précision thématique de classifications GEOBIA d'une image WorldView-2 / François Messner in Revue Française de Photogrammétrie et de Télédétection, n° 216 (février 2018)PermalinkSemantic enrichment of octree structured point clouds for multi‐story 3D pathfinding / Florian W. Fichtner in Transactions in GIS, vol 22 n° 1 (February 2018)PermalinkActive learning-based optimized training library generation for object-oriented image classification / Rajeswari Balasubramaniam in IEEE Transactions on geoscience and remote sensing, vol 56 n° 1 (January 2018)PermalinkAbove-bottom biomass retrieval of aquatic plants with regression models and SfM data acquired by a UAV platform – A case study in Wild Duck Lake Wetland, Beijing, China / Ran Jing in ISPRS Journal of photogrammetry and remote sensing, vol 134 (December 2017)PermalinkComplex-valued convolutional neural network and its application in polarimetric SAR image classification / Zhimian Zhang in IEEE Transactions on geoscience and remote sensing, vol 55 n° 12 (December 2017)PermalinkGlobal, dense multiscale reconstruction for a billion points / Benjamin Ummenhofer in International journal of computer vision, vol 125 n° 1-3 (December 2017)PermalinkMultilayer projective dictionary pair learning and sparse autoencoder for PolSAR image classification / Yanqiao Chen in IEEE Transactions on geoscience and remote sensing, vol 55 n° 12 (December 2017)PermalinkLearning a discriminative distance metric with label consistency for scene classification / Yuebin Wang in IEEE Transactions on geoscience and remote sensing, vol 55 n° 8 (August 2017)PermalinkLearning sensor-specific spatial-spectral features of hyperspectral images via convolutional neural networks / Shaohui Mei in IEEE Transactions on geoscience and remote sensing, vol 55 n° 8 (August 2017)PermalinkA morphologically preserved multi-resolution TIN surface modeling and visualization method for virtual globes / Xianwei Zheng in ISPRS Journal of photogrammetry and remote sensing, vol 129 (July 2017)Permalink