检索结果-内蒙古大学图书馆

Regularization of building boundaries in satellite images using adversarial and regularized losses

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zorzi, Stefano Fraundorfer, Friedrich Institute of Computer Graphics and Vision Graz University of Technology Austria

In this paper we present a method for building boundary refinement and regularization in satellite images using a fully convolutional neural network trained with a combination of adversarial and regularized losses. Compared to a pure Mask R-CNN model, the overall algorithm can achieve equivalent performance in terms of accuracy and completeness. However, unlike Mask R-CNN that produces irregular footprints, our framework generates regularized and visually pleasing building boundaries which are beneficial in many applications. Copyright © 2020, The Authors. All rights reserved.

关键词： Generative adversarial networks

Minimal Cases for Computing the Generalized Relative Pose using Affine Correspondences

学校读者我要写书评

暂无评论

Minimal Cases for Computing the Generalized Relative Pose us...

International Conference on computer vision (ICCV)

作者： Banglei Guan Ji Zhao Daniel Barath Friedrich Fraundorfer College of Aerospace Science and Engineering National University of Defense Technology China Computer Vision and Geometry Group ETH Zürich Institute for Computer Graphics and Vision Graz University of Technology Austria Remote Sensing Technology Institute German Aerospace Center Germany

ISBN: (纸本)9781665428132

We propose three novel solvers for estimating the relative pose of a multi-camera system from affine correspondences (ACs). A new constraint is derived interpreting the relationship of ACs and the generalized camera model. Using the constraint, we demonstrate efficient solvers for two types of motions assumed. Considering that the cameras undergo planar motion, we propose a minimal solution using a single AC and a solver with two ACs to overcome the degenerate case. Also, we propose a minimal solution using two ACs with known vertical direction, e.g., from an IMU. Since the proposed methods require significantly fewer correspondences than state-of-the-art algorithms, they can be efficiently used within RANSAC for outlier removal and initial motion estimation. The solvers are tested both on synthetic data and on real-world scenes from the KITTI odometry benchmark. It is shown that the accuracy of the estimated poses is superior to the state-of-the-art techniques.

关键词： computer vision Motion estimation Computational modeling Pose estimation Benchmark testing Cameras Image sequences

Ranking of attributes—comparative study based on data from stylometric domain

学校读者我要写书评

暂无评论

Procedia computer Science 2022年 207卷 2737-2746页

作者： Beata Zielosko Urszula Stańczyk Krzysztof Żabiński Institute of Computer Science University of Silesia in Katowice Będzińska 39 41-200 Sosnowiec Poland Department of Graphics Computer Vision and Digital Systems Faculty of Automatic Control Electronics and Computer Science Silesian University of Technology Akademicka 2A 44-100 Gliwice Poland

The area of feature selection methods constantly expands along with the development of artificial intelligence domain, and has great impact on almost every field, whenever data is processed and explored. The paper presents research where a ranking method was proposed, inspired by an approach which comes from an algorithm for induction of decision rules. The ranking procedure was based on calculation of standard deviation for attributes, taking into account assigned class labels. This method was compared with another ranking mechanism, a modified version of popular Relief algorithm, with incorporating characteristics of variables by supervised discretisation. Comparison of obtained results included the aspect of knowledge representation as well as the perspective of the accuracy for constructed rule-based classifiers. The experiments were performed on datasets from stylometry domain, where authorship attribution was considered as a classification task, and stylometric descriptors as characteristic features defining writing styles of authors.

关键词： Feature selection Ranking of attributes Relief Standard deviation Decision rules Stylometry Supervised discretisation Authorship attribution

Inferring the 3D Standing Spine Posture from 2D Radiographs 23rd

学校读者我要写书评

暂无评论

Inferring the 3D Standing Spine Posture from 2D Radiographs

23rd International Conference on Medical Image Computing and computer-Assisted Intervention, MICCAI 2020

作者： Bayat, Amirhossein Sekuboyina, Anjany Paetzold, Johannes C. Payer, Christian Stern, Darko Urschler, Martin Kirschke, Jan S. Menze, Bjoern H. Department of Informatics Technical University of Munich Munich Germany Department of Neuroradiology Klinikum rechts der Isar Munich Germany Institute of Computer Graphics and Vision Graz University of Technology Graz Austria School of Computer Science University of Auckland Auckland New Zealand

ISBN: (纸本)9783030597245

The treatment of degenerative spinal disorders requires an understanding of the individual spinal anatomy and curvature in 3D. An upright spinal pose (i.e. standing) under natural weight bearing is crucial for such bio-mechanical analysis. 3D volumetric imaging modalities (e.g. CT and MRI) are performed in patients lying down. On the other hand, radiographs are captured in an upright pose, but result in 2D projections. This work aims to integrate the two realms, i.e. it combines the upright spinal curvature from radiographs with the 3D vertebral shape from CT imaging for synthesizing an upright 3D model of spine, loaded naturally. Specifically, we propose a novel neural network architecture working vertebra-wise, termed TransVert, which takes orthogonal 2D radiographs and infers the spines 3D posture. We validate our architecture on digitally reconstructed radiographs, achieving a 3D reconstruction Dice of 95.52 %, indicating an almost perfect 2D-to-3D domain translation. Deploying our model on clinical radiographs, we successfully synthesise full-3D, upright, patient-specific spine models for the first time. © 2020, Springer Nature Switzerland AG.

关键词： Magnetic resonance imaging

Machine-learned regularization and polygonization of building segmentation masks

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zorzi, Stefano Bittner, Ksenia Fraundorfer, Friedrich Institute of Computer Graphics and Vision Graz University of Technology Austria Germany

We propose a machine learning based approach for automatic regularization and polygonization of building segmentation masks. Taking an image as input, we first predict building segmentation maps exploiting generic fully convolutional network (FCN). A generative adversarial network (GAN) is then involved to perform a regularization of building boundaries to make them more realistic, i.e., having more rectilinear outlines which construct right angles if required. This is achieved through the interplay between the discriminator which gives a probability of input image being true and generator that learns from discriminator’s response to create more realistic images. Finally, we train the backbone convolutional neural network (CNN) which is adapted to predict sparse outcomes corresponding to building corners out of regularized building segmentation results. Experiments on three building segmentation datasets demonstrate that the proposed method is not only capable of obtaining accurate results, but also of producing visually pleasing building outlines parameterized as polygons. Copyright © 2020, The Authors. All rights reserved.

关键词： Buildings

Map-Repair: Deep cadastre maps alignment and temporal inconsistencies fix in satellite images

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zorzi, Stefano Bittner, Ksenia Fraundorfer, Friedrich Institute of Computer Graphics and Vision Graz University of Technology Graz Austria Wessling Germany

In the fast developing countries it is hard to trace new buildings construction or old structures destruction and, as a result, to keep the up-to-date cadastre maps. Moreover, due to the complexity of urban regions or inconsistency of data used for cadastre maps extraction, the errors in form of misalignment is a common problem. In this work, we propose an end-to-end deep learning approach which is able to solve inconsistencies between the input intensity image and the available building footprints by correcting label noises and, at the same time, misalignments if needed. The obtained results demonstrate the robustness of the proposed method to even severely misaligned examples that makes it potentially suitable for real applications, like OpenStreetMap correction. Copyright © 2020, The Authors. All rights reserved.

关键词： Remote sensing

Generalisation of Segmentation Using Generative Adversarial Networks

学校读者我要写书评

暂无评论

Generalisation of Segmentation Using Generative Adversarial ...

IEEE International Symposium on Biomedical Imaging

作者： André Ferreira Gijs Luijten Behrus Puladi Jens Kleesiek Victor Alves Jan Egger Institute for AI in Medicine (IKIM) University Hospital Essen (AöR) Essen Germany Center Algoritmi / LASI University of Minho Braga Portugal Computer Algorithms for Medicine Laboratory Graz Austria Institute of Medical Informatics University Hospital RWTH Aachen Aachen Germany Department of Oral and Maxillofacial Surgery University Hospital RWTH Aachen Aachen Germany Institute of Computer Graphics and Vision (ICG) Graz University of Technology Graz Austria Cancer Research Center Cologne Essen (CCCE) West German Cancer Center Essen (WTZ) University Hospital Essen (AöR) Essen Germany Partner Site Essen German Cancer Consortium (DKTK) Essen Germany Department of Physics TU Dortmund University Dortmund Germany Center for Virtual and Extended Reality in Medicine (ZvRM) University Hospital Essen Essen Germany

ISBN: (数字)9798350313338

ISBN: (纸本)9798350313345

State-of-the-art deep learning algorithms are easily biased and evaluated in misleading scenarios, especially in the medical context, where scenarios change rapidly and diseases develop quickly. The BraTS 2024 GoAT challenge aims to evaluate how brain tumour segmentation algorithms can adapt to different circumstances when these are not available for training. Our solution utilises state-of-the-art conditional generative adversarial networks to generate realistic new cases and train a segmentation algorithm that takes advantage of the convolutions and attention mechanisms. Our solution achieved a DSC value of 0.855, 0.863, 0.883 and an HD95 value of 24.83, 24.10 and 21.72 for the enhancing tumour, the tumour core and the whole tumour in the validation set, respectively.

关键词： Training Knowledge engineering Deep learning Generative adversarial networks Robustness Inference algorithms Tumors

IB-MVS: An iterative algorithm for deep multi-view stereo based on binary decisions

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Sormann, Christian Rossi, Mattia Kuhn, Andreas Fraundorfer, Friedrich Institute of Computer Graphics and Vision Graz University of Technology Austria Sony Europe B.V. R&D Center - Stuttgart Laboratory 1 Germany

We present a novel deep-learning-based method for Multi-View Stereo. Our method estimates high resolution and highly precise depth maps iteratively, by traversing the continuous space of feasible depth values at each pixel in a binary decision fashion. The decision process leverages a deep-network architecture: this computes a pixelwise binary mask that establishes whether each pixel actual depth is in front or behind its current iteration individual depth hypothesis. Moreover, in order to handle occluded regions, at each iteration the results from different source images are fused using pixelwise weights estimated by a second network. Thanks to the adopted binary decision strategy, which permits an efficient exploration of the depth space, our method can handle high resolution images without trading resolution and precision. This sets it apart from most alternative learning-based Multi-View Stereo methods, where the explicit discretization of the depth space requires the processing of large cost volumes. We compare our method with state-of-the-art Multi-View Stereo methods on the DTU, Tanks and Temples and the challenging ETH3D benchmarks and show competitive results. Copyright © 2021, The Authors. All rights reserved.

关键词： Iterative methods

A Novel Approach for User Demand-aware Data Center Construction and Service Consolidation 1

学校读者我要写书评

暂无评论

29th International Conference on Web Services, ICWS 2022 Held as Part of the Services Conference Federation, SCF 2022

作者： Lv, Yifei Guo, Kunyin Xia, Yunni Li, Yin Ma, Yong Li, Fan Zeng, Linchengxi Peng, Qinglan Long, Tingyan College of Computer Science Chongqing University Chongqing400030 China Guangzhou Institute of Software Application Technology Guangzhou510000 China School of Computers and Information Jiangxi Normal University Nanchang330022 China Key Laboratory of Fundamental Synthetic Vision Graphics and Image Science for National Defense Sichuan University Chengdu610000 China Chongqing China School of Artificial Intelligence Henan University Kaifeng475001 China

ISBN: (数字)9783031235795

ISBN: (纸本)9783031235788

The recent intensifying computational demands from multinationals enterprises have motivated the magnification for large complicated cloud data centers (DCs) to handle business, monetary, Internet and commercial applications of different enterprises. A cloud data center encompasses thousands of physical server nodes arranged in racks along with network, storage, and other equipment that entails an extensive amount of power to process different processes and amenities required by business firms. More and more cloud data centers are turning for adapting to dynamics of user demands and reducing operational cost. Therefore, in this paper, we propose a user demand-aware (UDA) method for servers selection and a modified adaptive large neighbourhood search (MALNS) algorithm for dynamic service consolidation. Experiments based on real-world datasets demonstrate our approach outperformed conventional strategies in terms of multiple metrics. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Heuristic algorithms