Descripteur
Documents disponibles dans cette catégorie (4)
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 : Domain adaptation for urban scene segmentation Type de document : Thèse/HDR Auteurs : Antoine Saporta, Auteur ; Matthieu Cord, Directeur de thèse Editeur : Paris : Sorbonne Université Année de publication : 2022 Importance : 147 p. Format : 21 x 30 cm Note générale : bibliographie
Thèse de Doctorat de Sorbonne Université, spécialité InformatiqueLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] classification non dirigée
[Termes IGN] entropie
[Termes IGN] Mapillary
[Termes IGN] navigation autonome
[Termes IGN] réseau antagoniste génératif
[Termes IGN] scène urbaine
[Termes IGN] segmentation sémantique
[Termes IGN] vision par ordinateurIndex. décimale : THESE Thèses et HDR Résumé : (auteur) This thesis tackles some of the scientific locks of perception systems based on neural networks for autonomous vehicles. This dissertation discusses domain adaptation, a class of tools aiming at minimizing the need for labeled data. Domain adaptation allows generalization to so-called target data that share structures with the labeled so-called source data allowing supervision but nevertheless following a different statistical distribution. First, we study the introduction of privileged information in the source data, for instance, depth labels. The proposed strategy, BerMuDA, bases its domain adaptation on a multimodal representation obtained by bilinear fusion, modeling complex interactions between segmentation and depth. Next, we examine self-supervised learning strategies in domain adaptation, relying on selecting predictions on the unlabeled target data, serving as pseudo-labels. We propose two new selection criteria: first, an entropic criterion with ESL; then, with ConDA, using an estimate of the true class probability. Finally, the extension of adaptation scenarios to several target domains as well as in a continual learning framework is proposed. Two approaches are presented to extend traditional adversarial methods to multi-target domain adaptation: Multi-Dis. and MTKT. In a continual learning setting for which the target domains are discovered sequentially and without rehearsal, the proposed CTKT approach adapts MTKT to this new problem to tackle catastrophic forgetting. Note de contenu : 1- Introduction
2- Unsupervised domain adaptation
3- Leveraging priviledge information for unsupervised domain adaptation
4- Estimating and exploiting confident pseudo-labels for self-training
5- Adaptation to multiple domains
6- ConclusionNuméro de notice : 24079 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Thèse française Note de thèse : Thèse de Doctorat : Informatique : Sorbonne Université : 2022 Organisme de stage : Institut des Systèmes Intelligents et de Robotique DOI : sans En ligne : https://theses.hal.science/tel-03886201 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102213 Urban 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)
[article]
Titre : Urban infrastructure audit: an effective protocol to digitize signalized intersections by mining street view images Type de document : Article/Communication Auteurs : Xiao Li, Auteur ; Huan Ning, Auteur ; Xiao Huang, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 32 - 49 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] carrefour
[Termes IGN] cartographie urbaine
[Termes IGN] couche thématique
[Termes IGN] exploration d'images
[Termes IGN] feu de circulation
[Termes IGN] image Streetview
[Termes IGN] Mapillary
[Termes IGN] réseau routier
[Termes IGN] segmentation d'image
[Termes IGN] signalisation routièreRésumé : (auteur) Auditing and mapping traffic infrastructure is a crucial task in urban management. For example, signalized intersections play an essential role in transportation management; however, effectively identifying these intersections remains unsolved. Traditionally, signalized intersection data are manually collected through field audits or checking street view images (SVIs), which is time-consuming and labor-intensive. This study proposes an effective protocol to identify signalized intersections using road networks and SVIs. First, we propose a six-step geoprocessing model to generate an intersection feature layer from road networks. Second, we utilize up to three nearest SVIs to capture streetscapes at each intersection. Then, a deep learning-based image segmentation model is adopted to recognize traffic light-related pixels from each SVI. Last, we design a post-processing step to generate new features characterizing SVIs’ segmentation results at each intersection and build a decision tree model to determine the traffic control type. Results demonstrate that the proposed protocol can effectively identify signalized intersections with an overall accuracy of 97.05%. It also proves the effectiveness of SVIs for auditing urban infrastructures. This study can directly benefit transportation agencies by providing a ready-to-use smart audit and mapping solution for large-scale identification and mapping of signalized intersections. Numéro de notice : A2022-017 Affiliation des auteurs : non IGN Thématique : IMAGERIE/URBANISME Nature : Article DOI : 10.1080/15230406.2021.1992299 Date de publication en ligne : 16/11/2021 En ligne : https://doi.org/10.1080/15230406.2021.1992299 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99148
in Cartography and Geographic Information Science > vol 49 n° 1 (January 2022) . - pp 32 - 49[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 032-2022011 RAB Revue Centre de documentation En réserve L003 Disponible Towards generating network of bikeways from Mapillary data / Xuan Ding in Computers, Environment and Urban Systems, vol 88 (July 2021)
[article]
Titre : Towards generating network of bikeways from Mapillary data Type de document : Article/Communication Auteurs : Xuan Ding, Auteur ; Hongchao Fan, Auteur ; Jianya Gong, Auteur Année de publication : 2021 Article en page(s) : n° 101632 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique web
[Termes IGN] approche participative
[Termes IGN] cycliste
[Termes IGN] données localisées des bénévoles
[Termes IGN] gestion des itinéraires
[Termes IGN] Mapillary
[Termes IGN] OpenStreetMap
[Termes IGN] Suède
[Termes IGN] système d'information géographiqueRésumé : (auteur) Nowadays, biking is flourishing in many Western cities. While many roads are used for both cars and bicycles, buffered bike lanes are marked for the safety of cyclists. In many cities, segregated paths are built up to have physical separation from motor vehicles. These types of biking ways are regarded as attributes in geographic information system (GIS) data. This information is required and important in the service of route planning, as cyclists may prefer certain types of bikeways. This paper presents a framework for generating networks of bikeways with attribute information from the data collected on the collaborative street view data platform Mapillary. The framework consists of two layers: The first layer focuses on constructing a bikeway road network using Global Positioning System (GPS) information of Mapillary images. Mapillary sequences are classified into walking, cycling, driving (ordinary road), and driving (motorway) trajectories based on the transportation mode with a trained XGBoost classifier. The bikeway road network is then extracted from cycling and driving (ordinary road) trajectories using a raster-based method. The second layer focuses on extracting attribute information from Mapillary images. Cycling-specific information (i.e., bicycle signs/markings) is extracted using a two-stage detection and classification model. A series of quantitative evaluations based on a case study demonstrated the ability and potential of the framework for extracting bikeway road information to enrich the existing OSM cycling road data. Numéro de notice : A2021-432 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article DOI : 10.1016/j.compenvurbsys.2021.101632 Date de publication en ligne : 17/04/2021 En ligne : https://doi.org/10.1016/j.compenvurbsys.2021.101632 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97798
in Computers, Environment and Urban Systems > vol 88 (July 2021) . - n° 101632[article]A vélo au travers des Andes, pour OpenStreetMap / Anonyme in Géomatique expert, n° 126 (janvier - février 2019)
[article]
Titre : A vélo au travers des Andes, pour OpenStreetMap Type de document : Article/Communication Auteurs : Anonyme, Auteur Année de publication : 2019 Article en page(s) : pp 32 - 36 Langues : Français (fre) Descripteur : [Vedettes matières IGN] Géomatique web
[Termes IGN] bicyclette
[Termes IGN] collecte de données
[Termes IGN] contributeur
[Termes IGN] Cordillère des Andes
[Termes IGN] données localisées des bénévoles
[Termes IGN] image terrestre
[Termes IGN] Mapillary
[Termes IGN] OpenStreetMapRésumé : (Auteur) Alban Vivert a réalisé l’été dernier un voyage très particulier : avec sa simple bicyclette, baptisée la « poderosa » en hommage à Che Guevara, il a, tout comme lui, entrepris de parcourir une partie de l’Amérique latine. Mais avec un objectif différent : collecter le plus de photographies et d’informations géographiques possibles pendant son voyage, et les reverser sur Mapillary et Open Street Map. Numéro de notice : A2019-297 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article DOI : sans Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93263
in Géomatique expert > n° 126 (janvier - février 2019) . - pp 32 - 36[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité IFN-001-P002120 PER Revue Nogent-sur-Vernisson Salle périodiques Exclu du prêt