Centre de documentation
scientifique

Attention mechanisms in computer vision: A survey / Meng-Hao Guo in Computational Visual Media, vol 8 n° 3 (September 2022)  

Public [article]  inComputational Visual Media > vol 8 n° 3 (September 2022) . - pp 331 - 368 Titre : Attention mechanisms in computer vision: A survey Type de document : Article/Communication Auteurs : Meng-Hao Guo, Auteur ; Tian-Xing Xu, Auteur ; Jiang-Jiang Liu, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 331 - 368 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] attention (apprentissage automatique) [Termes IGN] saillance [Termes IGN] scène Résumé : (auteur) Humans can naturally and effectively find salient regions in complex scenes. Motivated by this observation, attention mechanisms were introduced into computer vision with the aim of imitating this aspect of the human visual system. Such an attention mechanism can be regarded as a dynamic weight adjustment process based on features of the input image. Attention mechanisms have achieved great success in many visual tasks, including image classification, object detection, semantic segmentation, video understanding, image generation, 3D vision, multimodal tasks, and self-supervised learning. In this survey, we provide a comprehensive review of various attention mechanisms in computer vision and categorize them according to approach, such as channel attention, spatial attention, temporal attention, and branch attention; a related repository https://github.com/MenghaoGuo/Awesome-Vision-Attentions is dedicated to collecting related work. We also suggest future directions for attention mechanism research. Numéro de notice : A2022-329 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Article DOI : 10.1007/s41095-022-0271-y Date de publication en ligne : 15/03/2022 En ligne : https://doi.org/10.1007/s41095-022-0271-y Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100601 [article]

Adaptive transfer of color from images to maps and visualizations / Mingguang Wu in Cartography and Geographic Information Science, Vol 49 n° 4 (July 2022)  

Public [article]  inCartography and Geographic Information Science > Vol 49 n° 4 (July 2022) . - pp 289 - 312 Titre : Adaptive transfer of color from images to maps and visualizations Type de document : Article/Communication Auteurs : Mingguang Wu, Auteur ; Yanjie Sun, Auteur ; Yaqian Li, Auteur Année de publication : 2022 Article en page(s) : pp 289 - 312 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] amélioration des couleurs [Termes IGN] couleur (rédaction cartographique) [Termes IGN] données vectorielles [Termes IGN] esthétique cartographique [Termes IGN] orthoimage couleur [Termes IGN] relation sémantique [Termes IGN] saillance [Termes IGN] visualisation cartographique [Vedettes matières IGN] Géovisualisation Résumé : (auteur) Because crafting attractive and effective colors from scratch is a high-effort and time-consuming process in map and visualization design, transferring color from an inspiration source to maps and visualizations is a promising technique for both novices and experts. To date, existing image-to-image color transfer methods suffer from ambiguities and inconsistencies; no computational approach is available to transfer color from arbitrary images to vector maps. To fill this gap, we propose a computational method that transfers color from arbitrary images to a vector map. First, we classify reference images into regions with measures of saliency. Second, we quantify the communicative quality and esthetics of colors in maps; we then transform the problem of color transfer into a dual-objective, multiple-constraint optimization problem. We also present a solution method that can create a series of optimal color suggestions and generate a communicative quality-esthetic compromise solution. We compare our method with an image-to-image method based on two sample maps and six reference images. The results indicate that our method is adaptive to mapping scales, themes, and regions. The evaluation also provides preliminary evidence that our method can achieve better communicative quality and harmony. Numéro de notice : A2022-478 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/15230406.2021.1982009 Date de publication en ligne : 10/11/2021 En ligne : https://doi.org/10.1080/15230406.2021.1982009 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100826 [article]

Summarizing large scale 3D mesh for urban navigation / Imeen Ben Salah in Robotics and autonomous systems, vol 152 (June 2022)  

Public [article]  inRobotics and autonomous systems > vol 152 (June 2022) . - n° 104037 Titre : Summarizing large scale 3D mesh for urban navigation Type de document : Article/Communication Auteurs : Imeen Ben Salah, Auteur ; Sébastien Kramm, Auteur ; Cédric Demonceaux, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 104037 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] algorithme ICP [Termes IGN] carte en 3D [Termes IGN] données lidar [Termes IGN] entropie [Termes IGN] image hémisphérique [Termes IGN] image RVB [Termes IGN] information sémantique [Termes IGN] localisation basée vision [Termes IGN] maillage [Termes IGN] navigation autonome [Termes IGN] précision géométrique (imagerie) [Termes IGN] précision radiométrique [Termes IGN] profondeur [Termes IGN] Rouen [Termes IGN] saillance [Termes IGN] zone urbaine Résumé : (auteur) Cameras have become increasingly common in vehicles, smartphones, and advanced driver assistance systems. The areas of application of these cameras in the world of intelligent transportation systems are becoming more and more varied: pedestrian detection, line crossing detection, navigation, …A major area of research currently focuses on mapping that is essential for localization and navigation. However, this step generates an important problem of memory management. Indeed, the memory space required to accommodate the map of a small city is measured in tens gigabytes. In addition, several providers today are competing to produce High-Definition (HD) maps. These maps offer a rich and detailed representation of the environment for highly accurate localization. However, they require a large storage capacity and high transmission and update costs. To overcome these problems, we propose a solution to summarize this type of map by reducing the size while maintaining the relevance of the data for navigation based on vision only. The summary consists in a set of spherical images augmented by depth and semantic information and allowing to keep the same level of visibility in every directions. These spheres are used as landmarks to offer guidance information to a distant agent. They then have to guarantee, at a lower cost, a good level of precision and speed during navigation. Some experiments on real data demonstrate the feasibility for obtaining a summarized map while maintaining a localization with interesting performances. Numéro de notice : A2022-290 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1016/j.robot.2022.104037 Date de publication en ligne : 03/02/2022 En ligne : https://doi.org/10.1016/j.robot.2022.104037 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100335 [article]

3D building model simplification method considering both model mesh and building structure / Jiangfeng She in Transactions in GIS, vol 26 n° 3 (May 2022)  

Public [article]  inTransactions in GIS > vol 26 n° 3 (May 2022) . - pp 1182 - 1203 Titre : 3D building model simplification method considering both model mesh and building structure Type de document : Article/Communication Auteurs : Jiangfeng She, Auteur ; Bo Chen, Auteur ; Junzhong Tan, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 1182 - 1203 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique [Termes IGN] contour [Termes IGN] contrainte géométrique [Termes IGN] empreinte [Termes IGN] maillage [Termes IGN] maillage par triangles [Termes IGN] modélisation 3D du bâti BIM [Termes IGN] saillance [Termes IGN] simplification de maillage [Termes IGN] simplification de surface Résumé : (auteur) The simplification of three-dimensional (3D) building models to improve rendering efficiency has gained widespread attention. To maintain the model's overall appearance features while increasing the simplification rate, we propose a novel 3D building simplification method that considers both the model mesh and building structure. The method divides a 3D building into a primary structure and subsidiary structures. It then organizes these structures using StructureTree, a multi-way tree. The structures are organized according to the dependency relationships between building structures. When simplifying a building, the decision whether to simplify the mesh or remove the subsidiary structure in the leaf node of the StructureTree depends on the volume change caused by the edge collapse and the visual saliency of the removed structure. The experimental results show that our method exhibits a better simplification effect than the traditional simplification method, and the proposed method can achieve a high simplification rate while maintaining the simplification quality. Furthermore, the results of some spatial analyses based on the highly simplified building model are consistent with those of the original model. Numéro de notice : A2022-464 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article DOI : https://doi.org/10.1111/tgis.12907 Date de publication en ligne : 14/02/2022 En ligne : https://doi.org/10.1111/tgis.12907 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100792 [article]

Unsupervised multi-view CNN for salient view selection and 3D interest point detection / Ran Song in International journal of computer vision, vol 130 n° 5 (May 2022)  

Public [article]  inInternational journal of computer vision > vol 130 n° 5 (May 2022) . - pp 1210 - 1227 Titre : Unsupervised multi-view CNN for salient view selection and 3D interest point detection Type de document : Article/Communication Auteurs : Ran Song, Auteur ; Wei Zhang, Auteur ; Yitian Zhao, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 1210 - 1227 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] classification non dirigée [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] détection d'objet [Termes IGN] objet 3D [Termes IGN] point d'intérêt [Termes IGN] saillance Résumé : (auteur) We present an unsupervised 3D deep learning framework based on a ubiquitously true proposition named by us view-object consistency as it states that a 3D object and its projected 2D views always belong to the same object class. To validate its effectiveness, we design a multi-view CNN instantiating it for salient view selection and interest point detection of 3D objects, which quintessentially cannot be handled by supervised learning due to the difficulty of collecting sufficient and consistent training data. Our unsupervised multi-view CNN, namely UMVCNN, branches off two channels which encode the knowledge within each 2D view and the 3D object respectively and also exploits both intra-view and inter-view knowledge of the object. It ends with a new loss layer which formulates the view-object consistency by impelling the two channels to generate consistent classification outcomes. The UMVCNN is then integrated with a global distinction adjustment scheme to incorporate global cues into salient view selection. We evaluate our method for salient view section both qualitatively and quantitatively, demonstrating its superiority over several state-of-the-art methods. In addition, we showcase that our method can be used to select salient views of 3D scenes containing multiple objects. We also develop a method based on the UMVCNN for 3D interest point detection and conduct comparative evaluations on a publicly available benchmark, which shows that the UMVCNN is amenable to different 3D shape understanding tasks. Numéro de notice : A2022-415 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1007/s11263-022-01592-x Date de publication en ligne : 16/03/2022 En ligne : https://doi.org/10.1007/s11263-022-01592-x Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=100771 [article]

Visual vs internal attention mechanisms in deep neural networks for image classification and object detection / Abraham Montoya Obeso in Pattern recognition, vol 123 (March 2022)  

Permalink

Generating geographical location descriptions with spatial templates: a salient toponym driven approach / Mark M. Hall in International journal of geographical information science IJGIS, vol 36 n° 1 (January 2022)  

Permalink

ComNet: combinational neural network for object detection in UAV-borne thermal images / Minglei Li in IEEE Transactions on geoscience and remote sensing, vol 59 n° 8 (August 2021)  

Permalink

CNN-based RGB-D salient object detection: Learn, select, and fuse / Hao Chen in International journal of computer vision, vol 129 n° 7 (July 2021)  

Permalink

Semantic-aware label placement for augmented reality in street view / Jianqing Jia in The Visual Computer, vol 37 n° 7 (July 2021)  

Permalink

Multi-level progressive parallel attention guided salient object detection for RGB-D images / Zhengyi Liu in The Visual Computer, vol 37 n° 3 (March 2021)  

Permalink

Perception de scène par un système multi-capteurs, application à la navigation dans des environnements d'intérieur structuré / Marwa Chakroun (2021)  

Permalink

Visualization of 3D property data and assessment of the impact of rendering attributes / Stefan Seipel in Journal of Geovisualization and Spatial Analysis, vol 4 n° 2 (December 2020)  

Permalink

A novel deep network and aggregation model for saliency detection / Ye Liang in The Visual Computer, vol 36 n° 9 (September 2020)  

Permalink

Comparing the roles of landmark visual salience and semantic salience in visual guidance during indoor wayfinding / Weihua Dong in Cartography and Geographic Information Science, vol 47 n° 3 (May 2020)  

Permalink