Centre de documentation
scientifique

Choosing an appropriate training set size when using existing data to train neural networks for land cover segmentation / Huan Ning in Annals of GIS, vol 26 n° 4 (October 2020)  

Public [article]  inAnnals of GIS > vol 26 n° 4 (October 2020) . - pp 329 - 342 Titre : Choosing an appropriate training set size when using existing data to train neural networks for land cover segmentation Type de document : Article/Communication Auteurs : Huan Ning, Auteur ; Zhenlong Li, Auteur ; Cuizhen Wang, Auteur ; et al., Auteur Année de publication : 2020 Article en page(s) : pp 329 - 342 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] apprentissage profond [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] contour [Termes IGN] extraction de traits caractéristiques [Termes IGN] jeu de données [Termes IGN] Kiangsi (Chine) [Termes IGN] occupation du sol [Termes IGN] segmentation d'image [Termes IGN] segmentation sémantique [Termes IGN] taille du jeu de données Résumé : (auteur) Land cover data is an inventory of objects on the Earth’s surface, which is often derived from remotely sensed imagery. Deep Convolutional Neural Network (DCNN) is a competitive method in image semantic segmentation. Some scholars argue that the inadequacy of training set is an obstacle when applying DCNNs in remote sensing image segmentation. While existing land cover data can be converted to large training sets, the size of training data set needs to be carefully considered. In this paper, we used different portions of a high-resolution land cover map to produce different sizes of training sets to train DCNNs (SegNet and U-Net) and then quantitatively evaluated the impact of training set size on the performance of the trained DCNN. We also introduced a new metric, Edge-ratio, to assess the performance of DCNN in maintaining the boundary of land cover objects. Based on the experiments, we document the relationship between the segmentation accuracy and the size of the training set, as well as the nonstationary accuracies among different land cover types. The findings of this paper can be used to effectively tailor the existing land cover data to training sets, and thus accelerate the assessment and employment of deep learning techniques for high-resolution land cover map extraction. Numéro de notice : A2020-800 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/19475683.2020.1803402 Date de publication en ligne : 10/08/2020 En ligne : https://doi.org/10.1080/19475683.2020.1803402 Format de la ressource électronique : url article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=96723 [article]

Interactions between hierarchical learning and visual system modeling : image classification on small datasets / Thalita Firmo Drumond (2020)  

Public Titre : Interactions between hierarchical learning and visual system modeling : image classification on small datasets Type de document : Thèse/HDR Auteurs : Thalita Firmo Drumond, Auteur ; Frédéric Alexandre, Directeur de thèse ; Thierry Viéville, Directeur de thèse Editeur : Bordeaux : Université de Bordeaux Année de publication : 2020 Importance : 195 p. Format : 21 x 30 cm Note générale : bibliographie Thèse présentée pour obtenir le grade de Docteur de l'Université de Bordeaux, Spécialité Informatique Langues : Français (fre) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] classification par réseau neuronal récurrent [Termes IGN] classification semi-dirigée [Termes IGN] corpus [Termes IGN] échantillonnage de données [Termes IGN] processus de hiérarchisation analytique [Termes IGN] reconnaissance d'objets [Termes IGN] taille du jeu de données [Termes IGN] vision par ordinateur Index. décimale : THESE Thèses et HDR Résumé : (auteur) Deep convolutional neural networks (DCNN) have recently protagonized a revolution in large-scale object recognition. They have changed the usual computer vision practices of hand-engineered features, with their ability to hierarchically learn representative features from data with a pertinent classifier. Together with hardware advances, they have made it possible to effectively exploit the ever-growing amounts of image data gathered online. However, in specific domains like healthcare and industrial applications, data is much less abundant, and expert labeling costs higher than those of general purpose image datasets. This scarcity scenario leads to this thesis' core question: can these limited-data domains profit from the advantages of DCNNs for image classification? This question has been addressed throughout this work, based on an extensive study of literature, divided in two main parts, followed by proposal of original models and mechanisms.The first part reviews object recognition from an interdisciplinary double-viewpoint. First, it resorts to understanding the function of vision from a biological stance, comparing and contrasting to DCNN models in terms of structure, function and capabilities. Second, a state-of-the-art review is established aiming to identify the main architectural categories and innovations in modern day DCNNs. This interdisciplinary basis fosters the identification of potential mechanisms - inspired both from biological and artificial structures — that could improve image recognition under difficult situations. Recurrent processing is a clear example: while not completely absent from the "deep vision" literature, it has mostly been applied to videos — due to their inherently sequential nature. From biology however it is clear such processing plays a role in refining our perception of a still scene. This theme is further explored through a dedicated literature review focused on recurrent convolutional architectures used in image classification.The second part carries on in the spirit of improving DCNNs, this time focusing more specifically on our central question: deep learning over small datasets. First, the work proposes a more detailed and precise discussion of the small sample problem and its relation to learning hierarchical features with deep models. This discussion is followed up by a structured view of the field, organizing and discussing the different possible paths towards adapting deep models to limited data settings. Rather than a raw listing, this review work aims to make sense out of the myriad of approaches in the field, grouping methods with similar intent or mechanism of action, in order to guide the development of custom solutions for small-data applications. Second, this study is complemented by an experimental analysis, exploring small data learning with the proposition of original models and mechanisms (previously published as a journal paper).In conclusion, it is possible to apply deep learning to small datasets and obtain good results, if done in a thoughtful fashion. On the data path, one shall try gather more information from additional related data sources if available. On the complexity path, architecture and training methods can be calibrated in order to profit the most from any available domain-specific side-information. Proposals concerning both of these paths get discussed in detail throughout this document. Overall, while there are multiple ways of reducing the complexity of deep learning with small data samples, there is no universal solution. Each method has its own drawbacks and practical difficulties and needs to be tailored specifically to the target perceptual task at hand. Note de contenu : 1- Introduction I- Object recognition with deep convolutional neural networks 2- Convolutional neural networks and visual system modeling 3- Feedforward CNN architectures for object recognition 4- Recurrent and feedback CNN architectures for object recognition II- Image classification on small datasets 5- A review of strategies to use deep learning under limited data 6- Analysis of DCNN applied to small sample learning using data prototypes 7 Conclusion Numéro de notice : 28312 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse française Note de thèse : thèse de Doctorat : Informatique : Bordeaux : 2020 Organisme de stage : Laboratoire bordelais de recherche en informatique DOI : sans En ligne : https://tel.archives-ouvertes.fr/tel-03129189v2/document Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=98233

Process BIM : Une chaîne de traitements pour le tel que construit / Tania Landes in Géomètre, n° 2146 (avril 2017)

Public [article]  inGéomètre > n° 2146 (avril 2017) . - pp 46 - 47 Titre : Process BIM : Une chaîne de traitements pour le tel que construit Type de document : Article/Communication Auteurs : Tania Landes, Auteur ; Pierre Grussenmeyer, Auteur Année de publication : 2017 Conférence : Colloque 2017, Le BIM : aux confluences de la technique et du droit 02/02/2017 Le Mans France Article en page(s) : pp 46 - 47 Langues : Français (fre) Descripteur : [Vedettes matières IGN] Lasergrammétrie [Termes IGN] chaîne de traitement [Termes IGN] extraction de données [Termes IGN] maquette numérique [Termes IGN] métier [Termes IGN] modélisation 3D du bâti BIM [Termes IGN] semis de points [Termes IGN] taille du jeu de données [Termes IGN] traitement de semis de points Résumé : (auteur) Le 30 janvier, Hélène Macher, ingénieur géomètre topographe diplômée de l'INSA Strasbourg, a soutenu sa thèse intitulée "Du nuage de points à la maquette numérique de bâtiment : reconstruction 3D semi-automatique de bâtiments existants". Numéro de notice : A2017-181 Affiliation des auteurs : non IGN Thématique : IMAGERIE/URBANISME Nature : Article DOI : sans Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=84736 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
063-2017041	RAB	Revue	Centre de documentation	En réserve L003	Disponible

Training set size, scale, and features in Geographic Object-Based Image Analysis of very high resolution unmanned aerial vehicle imagery / Lei Ma in ISPRS Journal of photogrammetry and remote sensing, vol 102 (April 2015)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 102 (April 2015) . - pp 14 - 27 Titre : Training set size, scale, and features in Geographic Object-Based Image Analysis of very high resolution unmanned aerial vehicle imagery Type de document : Article/Communication Auteurs : Lei Ma, Auteur ; Liang Cheng, Auteur ; Manchung Li, Auteur ; Yongxue Liu, Auteur ; Xiaoxue Ma, Auteur Année de publication : 2015 Article en page(s) : pp 14 - 27 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] analyse d'image orientée objet [Termes IGN] classification orientée objet [Termes IGN] drone [Termes IGN] échelle de prise de vue [Termes IGN] image à ultra haute résolution [Termes IGN] image aérienne [Termes IGN] image optique [Termes IGN] taille du jeu de données Résumé : (auteur) Unmanned Aerial Vehicle (UAV) has been used increasingly for natural resource applications in recent years due to their greater availability and the miniaturization of sensors. In addition, Geographic Object-Based Image Analysis (GEOBIA) has received more attention as a novel paradigm for remote sensing earth observation data. However, GEOBIA generates some new problems compared with pixel-based methods. In this study, we developed a strategy for the semi-automatic optimization of object-based classification, which involves an area-based accuracy assessment that analyzes the relationship between scale and the training set size. We found that the Overall Accuracy (OA) increased as the training set ratio (proportion of the segmented objects used for training) increased when the Segmentation Scale Parameter (SSP) was fixed. The OA increased more slowly as the training set ratio became larger and a similar rule was obtained according to the pixel-based image analysis. The OA decreased as the SSP increased when the training set ratio was fixed. Consequently, the SSP should not be too large during classification using a small training set ratio. By contrast, a large training set ratio is required if classification is performed using a high SSP. In addition, we suggest that the optimal SSP for each class has a high positive correlation with the mean area obtained by manual interpretation, which can be summarized by a linear correlation equation. We expect that these results will be applicable to UAV imagery classification to determine the optimal SSP for each class. Numéro de notice : A2015-692 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2014.12.026 En ligne : https://doi.org/10.1016/j.isprsjprs.2014.12.026 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=78323 [article]