Détail de l'éditeur
ArXiv - Université Cornell
localisé à :
Ithaca
|
Documents disponibles chez cet éditeur (27)
Ajouter le résultat dans votre panier
Visionner les documents numériques
Affiner la recherche Interroger des sources externesBuyTheDips : PathLoss for improved topology-preserving deep learning-based image segmentation / Minh On Vu Ngoc (2022)
Titre : BuyTheDips : PathLoss for improved topology-preserving deep learning-based image segmentation Type de document : Article/Communication Auteurs : Minh On Vu Ngoc, Auteur ; Yizi Chen Editeur : Ithaca [New York - Etats-Unis] : ArXiv - Université Cornell Année de publication : 2022 Projets : SODUCO / Perret, Julien Importance : 13 p. Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] chemin le plus court, algorithme du
[Termes IGN] fonction de perte
[Termes IGN] image numérique
[Termes IGN] proximité sémantique
[Termes IGN] segmentation d'imageRésumé : (auteur) Capturing the global topology of an image is essential for proposing an accurate segmentation of its domain. However, most of existing segmentation methods do not preserve the initial topology of the given input, which is detrimental for numerous downstream object-based tasks. This is all the more true for deep learning models which most work at local scales. In this paper, we propose a new topology-preserving deep image segmentation method which relies on a new leakage loss: the Pathloss. Our method is an extension of the BALoss [1], in which we want to improve the leakage detection for better recovering the closeness property of the image segmentation. This loss allows us to correctly localize and fix the critical points (a leakage in the boundaries) that could occur in the predictions, and is based on a shortest-path search algorithm. This way, loss minimization enforces connectivity only where it is necessary and finally provides a good localization of the boundaries of the objects in the image. Moreover, according to our research, our Pathloss learns to preserve stronger elongated structure compared to methods without using topology-preserving loss. Training with our topological loss function, our method outperforms state-of-the-art topology-aware methods on two representative datasets of different natures: Electron Microscopy and Historical Map. Numéro de notice : P2022-005 Affiliation des auteurs : UGE-LASTIG+Ext (2020- ) Thématique : IMAGERIE/INFORMATIQUE Nature : Preprint nature-HAL : Préprint DOI : 10.48550/arXiv.2207.11446 En ligne : https://doi.org/10.48550/arXiv.2207.11446 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101338
Titre : Deep surface reconstruction from point clouds with visibility information Type de document : Article/Communication Auteurs : Raphaël Sulzer Editeur : Ithaca [New York - Etats-Unis] : ArXiv - Université Cornell Année de publication : 2022 Projets : BIOM / Vallet, Bruno Conférence : ICPR 2022, 26th International Conference on Pattern Recognition 21/08/2022 25/08/2022 Montréal Québec - Canada Proceedings IEEE Importance : 13 p. Format : 21 x 30 cm Note générale : bibliographie
https://doi.org/10.48550/arXiv.2202.01810 sur ArXivLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] reconstruction d'objet
[Termes IGN] semis de points
[Termes IGN] visibilitéRésumé : (auteur) Most current neural networks for reconstructing surfaces from point clouds ignore sensor poses and only operate on raw point locations. Sensor visibility, however, holds meaningful information regarding space occupancy and surface orientation. In this paper, we present two simple ways to augment raw point clouds with visibility information, so it can directly be leveraged by surface reconstruction networks with minimal adaptation. Our proposed modifications consistently improve the accuracy of generated surfaces as well as the generalization ability of the networks to unseen shape domains. Numéro de notice : C2022-048 Affiliation des auteurs : UGE-LASTIG+Ext (2020- ) Autre URL associée : vers HAL Thématique : IMAGERIE Nature : Communication nature-HAL : ComAvecCL&ActesPubliésIntl DOI : 10.48550/arXiv.2202.01810 Date de publication en ligne : 03/02/2022 En ligne : https://doi.org/10.1109/ICPR56361.2022.9956560 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99811
Titre : GENESIS: Co-location of Geodetic Techniques in Space Type de document : Article/Communication Auteurs : Pacôme Delva, Auteur ; Zuheir Altamimi Editeur : Ithaca [New York - Etats-Unis] : ArXiv - Université Cornell Année de publication : 2022 Projets : 1-Pas de projet / Vallet, Bruno Note générale : bibliographie
auteurs : Pacome Delva, Zuheir Altamimi, Alejandro Blazquez, Mathis Blossfeld, Johannes Böhm, Pascal Bonnefond, Jean-Paul Boy, Sean Bruinsma, Grzegorz Bury, Miltiadis Chatzinikos, Alexandre Couhert, Clement Courde, Rolf Dach, Veronique Dehant, Simone Dell’Agnello, Gunnar Elgered, Werner Enderle, Pierre Exertier, Susanne Glaser, Rudiger Haas, Wen Huang, Urs Hugentobler17, Adrian J¨aggi11, Ozgur Karatekin12, Frank G. Lemoine18, Christophe Le Poncin-Lafitte, Susanne Lunz, Benjamin Mannel, Flavien Mercier, Laurent Metivier, Benoıt Meyssignac, Jurgen Muller, Axel Nothnage, Felix Perosanz, Roelof Rietbroek, Markus Rothacher, Hakan Sert, Krzysztof Sosnica, Paride Testani, Javier Ventura-Traveset, Gilles
Wautelet, and Radoslaw ZajdeLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Systèmes de référence et réseaux
[Termes IGN] co-positionnement
[Termes IGN] état de l'art
[Termes IGN] International Terrestrial Reference Frame
[Termes IGN] mission spatiale
[Termes IGN] station de mesureRésumé : (auteur) Improving and homogenizing time and space reference systems on Earth and, more directly, realizing the Terrestrial Reference Frame (TRF) with an accuracy of 1mm and a long-term stability of 0.1mm/year are relevant for many scientific and societal endeavors. The knowledge of the TRF is fundamental for Earth and navigation sciences. For instance, quantifying sea level change strongly depends on an accurate determination of the geocenter motion but also of the positions of continental and island reference stations, as well as the ground stations of tracking networks. Also, numerous applications in geophysics require absolute millimeter precision from the reference frame, as for example monitoring tectonic motion or crustal deformation for predicting natural hazards. The TRF accuracy to be achieved represents the consensus of various authorities which has enunciated geodesy requirements for Earth sciences.
Today we are still far from these ambitious accuracy and stability goals for the realization of the TRF. However, a combination and co-location of all four space geodetic techniques on one satellite platform can significantly contribute to achieving these goals. This is the purpose of the GENESIS mission, proposed as a component of the FutureNAV program of the European Space Agency. The GENESIS platform will be a dynamic space geodetic observatory carrying all the geodetic instruments referenced to one another through carefully calibrated space ties. The co-location of the techniques in space will solve the inconsistencies and biases between the different geodetic techniques in order to reach the TRF accuracy and stability goals endorsed by the various international authorities and the scientific community. The purpose of this white paper is to review the state-of-the-art and explain the benefits of the GENESIS mission in Earth sciences, navigation sciences and metrology.Numéro de notice : P2022-007 Affiliation des auteurs : UMR IPGP-Géod+Ext (2020- ) Thématique : POSITIONNEMENT Nature : Preprint nature-HAL : Préprint DOI : 10.48550/arXiv.2209.15298 Date de publication en ligne : 30/09/2022 En ligne : https://doi.org/10.48550/arXiv.2209.15298 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101792
Titre : Learning multi-view aggregation in the wild for large-scale 3D semantic segmentation Type de document : Article/Communication Auteurs : Damien Robert Editeur : Ithaca [New York - Etats-Unis] : ArXiv - Université Cornell Année de publication : 2022 Projets : 3-projet - voir note / Vallet, Bruno Conférence : CVPR 2022 19/06/2022 24/06/2022 New Orleans Louisiane - Etats-Unis OA Proceedings Importance : pp 5575 - 5584 Note générale : bibliographie
This work was funded by ENGIE Lab CRIGEN and carried on in the LASTIG research unit of Universite Paris-Est. The authors wish to thank AI4GEO for sharing their computing resources.Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques
[Termes IGN] chaîne de traitement
[Termes IGN] données localisées 2D
[Termes IGN] données localisées 3D
[Termes IGN] données massives
[Termes IGN] pixel
[Termes IGN] segmentation sémantique
[Termes IGN] semis de pointsRésumé : (auteur) Recent works on 3D semantic segmentation propose to exploit the synergy between images and point clouds by processing each modality with a dedicated network and projecting learned 2D features onto 3D points. Merging large-scale point clouds and images raises several challenges, such as constructing a mapping between points and pixels, and aggregating features between multiple views. Current methods require mesh reconstruction or specialized sensors to recover occlusions, and use heuristics to select and aggregate available images. In contrast, we propose an end-to-end trainable multi-view aggregation model leveraging the viewing conditions of 3D points to merge features from images taken at arbitrary positions. Our method can combine standard 2D and 3D networks and outperforms both 3D models operating on colorized point clouds and hybrid 2D/3D networks without requiring colorization, meshing, or true depth maps. We set a new state-of-the-art for large-scale indoor/ outdoor semantic segmentation on S3DIS (74.7 mIoU 6-Fold) and on KITTI360 (58.3 mIoU). Our full pipeline is accessible at https: //github.com/drprojects/DeepViewAgg, and only requires raw 3D scans and a set of images and poses. Numéro de notice : C2022-006 Affiliation des auteurs : UGE-LASTIG (2020- ) Autre URL associée : vers CVF Thématique : IMAGERIE/INFORMATIQUE Nature : Communication nature-HAL : ComAvecCL&ActesPubliésIntl DOI : 10.48550/arXiv.2204.07548 Date de publication en ligne : 15/04/2022 En ligne : https://doi.org/10.48550/arXiv.2204.07548 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100490
Titre : Location retrieval using qualitative place signatures of visible landmarks Type de document : Article/Communication Auteurs : Lijun Wei Editeur : Ithaca [New York - Etats-Unis] : ArXiv - Université Cornell Année de publication : 2022 Projets : 1-Pas de projet / Vallet, Bruno Importance : 52 p. Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] descripteur
[Termes IGN] lieu
[Termes IGN] point de repère
[Termes IGN] reconnaissance d'objets
[Termes IGN] relation spatialeRésumé : (auteur) Location retrieval based on visual information is to retrieve the location of an agent (e.g. human, robot) or the area they see by comparing the observations with a certain form of representation of the environment. Existing methods generally require precise measurement and storage of the observed environment features, which may not always be robust due to the change of season, viewpoint, occlusion, etc. They are also challenging to scale up and may not be applicable for humans due to the lack of measuring/imaging devices. Considering that humans often use less precise but easily produced qualitative spatial language and high-level semantic landmarks when describing an environment, a qualitative location retrieval method is proposed in this work by describing locations/places using qualitative place signatures (QPS), defined as the perceived spatial relations between ordered pairs of co-visible landmarks from viewers' perspective. After dividing the space into place cells each with individual signatures attached, a coarse-to-fine location retrieval method is proposed to efficiently identify the possible location(s) of viewers based on their qualitative observations. The usability and effectiveness of the proposed method were evaluated using openly available landmark datasets, together with simulated observations by considering the possible perception error. Numéro de notice : P2022-009 Affiliation des auteurs : UGE-LASTIG+Ext (2020- ) Thématique : IMAGERIE Nature : Preprint nature-HAL : Préprint DOI : 10.48550/arXiv.2208.00783 Date de publication en ligne : 26/07/2022 En ligne : https://doi.org/10.48550/arXiv.2208.00783 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101879 PermalinkPermalinkPermalinkPermalinkPermalinkPermalinkPermalinkPermalinkPermalinkPermalink
