Centre dedocumentation
scientifique

Deep learning based 2D and 3D object detection and tracking on monocular video in the context of autonomous vehicles / Zhujun Xu (2022)  

Public Titre : Deep learning based 2D and 3D object detection and tracking on monocular video in the context of autonomous vehicles Type de document : Thèse/HDR Auteurs : Zhujun Xu, Auteur ; Eric Chaumette, Directeur de thèse ; Damien Vivet, Directeur de thèse Editeur : Toulouse : Université de Toulouse Année de publication : 2022 Importance : 136 p. Format : 21 x 30 cm Note générale : bibliographie Thèse en vue de l'obtention du Doctorat de l'Université de Toulouse, spécialité Informatique et Télécommunications Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] apprentissage semi-dirigé [Termes IGN] architecture de réseau [Termes IGN] détection d'objet [Termes IGN] échantillonnage de données [Termes IGN] objet 3D [Termes IGN] segmentation d'image [Termes IGN] véhicule automobile [Termes IGN] vidéo [Termes IGN] vision par ordinateur Index. décimale : THESE Thèses et HDR Résumé : (auteur) The objective of this thesis is to develop deep learning based 2D and 3D object detection and tracking methods on monocular video and apply them to the context of autonomous vehicles. Actually, when directly using still image detectors to process a video stream, the accuracy suffers from sampled image quality problems. Moreover, generating 3D annotations is time-consuming and expensive due to the data fusion and large numbers of frames. We therefore take advantage of the temporal information in videos such as the object consistency, to improve the performance. The methods should not introduce too much extra computational burden, since the autonomous vehicle demands a real-time performance.Multiple methods can be involved in different steps, for example, data preparation, network architecture and post-processing. First, we propose a post-processing method called heatmap propagation based on a one-stage detector CenterNet for video object detection. Our method propagates the previous reliable long-term detection in the form of heatmap to the upcoming frame. Then, to distinguish different objects of the same class, we propose a frame-to-frame network architecture for video instance segmentation by using the instance sequence queries. The tracking of instances is achieved without extra post-processing for data association. Finally, we propose a semi-supervised learning method to generate 3D annotations for 2D video object tracking dataset. This helps to enrich the training process for 3D object detection. Each of the three methods can be individually applied to leverage image detectors to video applications. We also propose two complete network structures to solve 2D and 3D object detection and tracking on monocular video. Note de contenu : 1- Introduction 2- Video object detection avec la heatmap propagation (propagation de carte de chaleur) 3- Video instance segmentation with instance sequence queries 4- Semi-supervised learning of monocular 3D object detection with 2D video tracking annotations 5- Conclusions and perspectives Numéro de notice : 24072 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse française Note de thèse : Thèse de Doctorat : Informatique et Télécommunications : Toulouse : 2022 DOI : sans En ligne : https://www.theses.fr/2022ESAE0019 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102136

The position of sound in audiovisual maps: an experimental study of performance in spatial memory / Nils Siepmann in Cartographica, vol 55 n° 2 (Summer 2020)  

Public [article]  inCartographica > vol 55 n° 2 (Summer 2020) . - pp 136 - 150 Titre : The position of sound in audiovisual maps: an experimental study of performance in spatial memory Type de document : Article/Communication Auteurs : Nils Siepmann, Auteur ; Dennis Edler, Auteur ; Julian Keil, Auteur ; et al., Auteur Année de publication : 2020 Article en page(s) : pp 136 - 150 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Cartographie numérique [Termes IGN] audiovisuel [Termes IGN] carte cognitive [Termes IGN] communication cartographique [Termes IGN] document sonore [Termes IGN] information géographique [Termes IGN] information sémantique [Termes IGN] mémoire [Termes IGN] multimedia Résumé : (auteur) Digital maps are known as reliable media for communicating spatial information. People use maps to make themselves familiar with new environments and to form cognitive representations of spatial configurations and additional semantic information that are coupled with locational information. Since the mid-1990s, cartographers have explored auditory media as cartographic elements to transfer spatial information. Among the established sound variants used in multimedia cartography, speech recordings are a popular auditory tool to enrich the visual dominance of maps. The impact of auditory elements on human spatial memory has hardly been investigated so far in cartography and spatial cognition. A recent study showed that spoken object names bound to visual location markers affect performance in memory of object locations. Map users tend to make significantly smaller spatial distortion errors in the recall of object locations if these locations are coupled with auditory semantic information (place names). The present study extends this approach by examining possible effects on sound position as cues for spatial memory performance. A monaural condition, where an auditory name is presented in a spatial location corresponding to the object location, is compared with a binaural condition (of no directional cue). The results show that a monaural communication additionally improves spatial memory performance. Interestingly, the semantic information bound to an object location appears to be the driving factor in improving this effect. Numéro de notice : A2020-441 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.3138/cart-2019-0008 Date de publication en ligne : 16/06/2020 En ligne : https://doi.org/10.3138/cart-2019-0008 Format de la ressource électronique : url article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=95499 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
031-2020021	SL	Revue	Centre de documentation	Revues en salle	Disponible

Audiovisual cartography : Established and new multimedia approaches to represent soundscapes / Dennis Edler in KN, Journal of Cartography and Geographic Information, vol 69 n° 1 (May 2019)  

Public [article]  inKN, Journal of Cartography and Geographic Information > vol 69 n° 1 (May 2019) . - pp 5 - 17 Titre : Audiovisual cartography : Established and new multimedia approaches to represent soundscapes Type de document : Article/Communication Auteurs : Dennis Edler, Auteur ; Olaf Kühne, Auteur ; Julian Keil, Auteur ; Frank Dickmann, Auteur Année de publication : 2019 Article en page(s) : pp 5 - 17 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Cartographie [Termes IGN] audiovisuel [Termes IGN] moteur de jeu [Termes IGN] multimedia [Termes IGN] paysage sonore [Termes IGN] réalité virtuelle Résumé : (Auteur) Since the mid-1990s, sound has been discussed and used in multimedia cartography. There are four main variants of auditory map elements that have been established in the theory and practice of audiovisual cartography, i.e. abstract sounds/abstract sound sequences, speech, music and, especially, audiorealistic sequences representing so-called “soundscapes”. In cartography, soundscapes are often addressed in large-scale representations. The term originates in multidisciplinary landscape research. Empirical findings of landscape research, and especially those of social constructivist landscape research, have shown the relevance of non-visual stimuli for people’s individual impressions and meanings of the experienced landscape. Amongst the non-visual landscape dimensions, the auditory dimension is the most prominent one. As 3D cartography offers new methods and techniques of designing and experiencing highly realistic, incl. photo- and audiorealistic landscape representation, this discipline becomes more and more interesting for simulating and presenting multisensory landscapes and for presenting the results of empirical findings in landscape research. After an introduction to traditional means of audiovisual cartography and the relevance of auditory stimuli for social constructivist approaches of landscape research, a modern software-based method is presented which highlights the opportunity to imbed 3D sound data representing a location’s soundscape into audiovisual 3D environments in Virtual Reality (VR). It is technically based on the cross-platform game engine Unity3D. Numéro de notice : A2019-177 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s42489-019-00004-4 Date de publication en ligne : 04/02/2019 En ligne : http://dx.doi.org/10.1007/s42489-019-00004-4 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92697 [article]

The simplicity of modern audiovisual web cartography : An example with the open-source JavaScript library leaflet.js / Dennis Edler in KN, Journal of Cartography and Geographic Information, vol 69 n° 1 (May 2019)  

Public [article]  inKN, Journal of Cartography and Geographic Information > vol 69 n° 1 (May 2019) . - pp 51 - 62 Titre : The simplicity of modern audiovisual web cartography : An example with the open-source JavaScript library leaflet.js Type de document : Article/Communication Auteurs : Dennis Edler, Auteur ; Mark Vetter, Auteur Année de publication : 2019 Article en page(s) : pp 51 - 62 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique web [Termes IGN] audiovisuel [Termes IGN] cartographie par internet [Termes IGN] format JSON [Termes IGN] Javascript (langage de script) [Termes IGN] Leaflet [Termes IGN] logiciel libre [Termes IGN] OpenStreetMap Résumé : (Auteur) Audio files entered cartography about 25 years ago. Since the mid-1990s, several examples of audiovisual maps have been created and published. These maps cover a diverse range of applications. In terms of the sound elements implemented, four characteristic variants were mainly used: abstract sounds/sound sequences, language recordings, and music and recordings/simulations of the real soundscape. To built audiovisual maps technically, several tools of animation software solutions were intensively explored. The rising importance of the World Wide Web also increased the demand for new and modern web-based approaches of multimedia cartography. One of the core technologies of the Internet is the programming language JavaScript (JS). The language is often associated with the so-called libraries which are applied programming packages pre-written for particular purposes. In cartography, one of the most popular and widely used JS libraries is leaflet.js. This open-source library is known for its simplicity and, meanwhile, it also supports a straightforward integration of multimedia content, such as audio files. It also teams up with HTML5 and CSS3. This paper intends to give a workflow focused on how to create individual audiovisual web mapping applications based on the contemporary options offered by leaflet.js. Numéro de notice : A2019-178 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s42489-019-00006-2 Date de publication en ligne : 12/02/2019 En ligne : https://doi.org/10.1007/s42489-019-00006-2 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92698 [article]

The sound of being there : Audiovisual cartography with immersive virtual environments / Florian Hruby in KN, Journal of Cartography and Geographic Information, vol 69 n° 1 (May 2019)  

Public [article]  inKN, Journal of Cartography and Geographic Information > vol 69 n° 1 (May 2019) . - pp 19 - 28 Titre : The sound of being there : Audiovisual cartography with immersive virtual environments Type de document : Article/Communication Auteurs : Florian Hruby, Auteur Année de publication : 2019 Article en page(s) : pp 19 - 28 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] audiovisuel [Termes IGN] environnement géographique virtuel [Termes IGN] immersion [Termes IGN] réalité virtuelle [Vedettes matières IGN] Géovisualisation Résumé : (Auteur) In recent years, immersion has become a frequently emphasized factor in the geovisualization research agenda. A principal reason for this growing interest is the assumption that immersive virtual environments (IVE) facilitate the formation of spatial presence, generally understood as the sense of “being there”. In a virtually mediated environment, the feeling of being there is of particular concern for cartographic ambitions, in terms of generating insights through geospatial representation. Current literature indicates that immersive VR systems stimulate the experience of spatial presence; however, this assumption is mainly based upon user studies in the visual communication channel. Moreover, research on IVE for geovisualization matters has to date been focused on visual-graphical rather than on auditive or even multisensory representations in virtual space. In this context, the present paper aims to evaluate the potential of audiovisual cartography with immersive virtual environments. Following a brief discussion of basic concepts, such as immersion, spatial presence and embodiment, we will integrate these aspects into a geovisualization immersion pipeline (GIP), as a framework with which to systematically link the technical and cognitive aspects of IVE. In the subsequent sections, we will examine this framework in the audio channel by analyzing how sound is implemented and perceived in GeoIVE. As we shall see, the positive effect of a combined audio-visual vs. exclusively visual presentation is supported by a series of user studies of sound effects, making audiovisual cartography with IVE a rich and worthwhile field of research. Numéro de notice : A2019-179 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1007/s42489-019-00003-5 Date de publication en ligne : 28/01/2019 En ligne : https://doi.org/10.1007/s42489-019-00003-5 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92699 [article]

Multiple cues-based active contours for target contour tracking under sophisticated background / Peng Lv in The Visual Computer, vol 33 n°9 (September 2017)  

Permalink

Cartographie et interprétation de l'environnement par drone / Martial Sanfourche in Revue Française de Photogrammétrie et de Télédétection, n° 213 - 214 (janvier - avril 2017)  

Permalink

Deck and cable dynamic testing of a single-span bridge using radar interferometry and videometry measurements / George Piniotis in Journal of applied geodesy, vol 10 n° 1 (March 2016)  

Permalink

Microwave unmixing with video segmentation for inferring broadleaf and needleleaf brightness temperatures and abundances from mixed forest observations / Lingjia Gu in IEEE Transactions on geoscience and remote sensing, vol 54 n° 1 (January 2016)  

Permalink

Spatio-temporal traffic video data archiving and retrieval system / Hang Yue in Geoinformatica, vol 20 n° 1 (January - March 2016)  

Permalink

Large-scale geo-tagged video indexing and queries / He Ma in Geoinformatica, vol 18 n° 4 (October 2014)  

Permalink

The Bulger case : A spatial story / Les Roberts in Cartographic journal (the), vol 51 n° 2 (May 2014)  

Permalink

A user study of a map-based slideshow editor / H. Fujita in Cartographica, vol 46 n° 2 (June 2011)  

Permalink

Création de cartes animées en vidéo haute définition / B. Saglio (2011)

Permalink

A new standard for analyzing geospatial intelligence, Geo-fused video / W. Smith in Geoinformatics, vol 14 n° 1 (01/01/2011)

Permalink