检索结果-内蒙古大学图书馆

HOC-Search: Efficient CAD Model and Pose Retrieval from RGB-D Scans

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Ainetter, Stefan Stekovic, Sinisa Fraundorfer, Friedrich Lepetit, Vincent Institute for Computer Graphics and Vision Graz University of Technology Graz Austria LIGM École des Ponts Univ Gustave Eiffel CNRS Marne-la-Vallée France

We present an automated and efficient approach for retrieving high-quality CAD models of objects and their poses in a scene captured by a moving RGB-D camera. We first investigate various objective functions to measure similarity between a candidate CAD object model and the available data, and the best objective function appears to be a"render-and-compare" method comparing depth and mask rendering. We thus introduce a fast-search method that approximates an exhaustive search based on this objective function for simultaneously retrieving the object category, a CAD model, and the pose of an object given an approximate 3D bounding box. This method involves a search tree that organizes the CAD models and object properties including object category and pose for fast retrieval and an algorithm inspired by Monte Carlo Tree Search, that efficiently searches this tree. We show that this method retrieves CAD models that fit the real objects very well, with a speed-up factor of 10x to 120x compared to exhaustive search. Copyright © 2023, The Authors. All rights reserved.

关键词： computer aided design

Moment-based opacity optimization

学校读者我要写书评

暂无评论

Moment-based opacity optimization

20th Eurographics Symposium on Parallel graphics and Visualization, EGPGV 2020

作者： Zeidan, M. Rapp, T. Peters, C. Dachsbacher, C. Computer Graphics Group Karlsruhe Institute of Technology Germany

ISBN: (纸本)9783038681076

Geometric structures such as points, lines, and surfaces play a vital role in scientific visualization. However, these visualizations frequently suffer from visual clutter that hinders the inspection of important features behind dense but less important features. In the past few years, geometric cluttering and occlusion avoidance has been addressed in scientific visualization with various approaches such as opacity optimization techniques. In this paper, we present a novel approach for opacity optimization based on recent state-of-the-art moment-based techniques for signal reconstruction. In contrast to truncated Fourier series, moment-based reconstructions of feature importance and optical depth along view rays are highly accurate for sparse regions but also plausible for densely covered regions. At the same time, moment-based methods do not suffer from ringing artifacts. Moreover, this representation enables fast evaluation and compact storage, which is crucial for per-pixel optimization especially for large geometric structures. We also present a fast screen space filtering approach for optimized opacities that works directly on moment buffers. This filtering approach is suitable for real-time visualization applications, while providing comparable quality to object space smoothing. Its implementation is independent of the type of geometry such that it is general and easy to integrate. We compare our technique to recent state of the art techniques for opacity optimization and apply it to real and synthetic data sets in various applications. © 2020 The Author(s) Eurographics Proceedings © 2020 The Eurographics Association.

关键词：

SAILOR: Scaling Anchors via Insights into Latent Object Representation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Malić, Dušan Fruhwirth-Reisinger, Christian Possegger, Horst Bischof, Horst Institute of Computer Graphics and Vision Graz University of Technology Austria Christian Doppler Laboratory for Embedded Machine Learning Austria

LiDAR 3D object detection models are inevitably biased towards their training dataset. The detector clearly exhibits this bias when employed on a target dataset, particularly towards object sizes. However, object sizes vary heavily between domains due to, for instance, different labeling policies or geographical locations. State-of-the-art unsupervised domain adaptation approaches outsource methods to overcome the object size bias. Mainstream size adaptation approaches exploit target domain statistics, contradicting the original unsupervised assumption. Our novel unsupervised anchor calibration method addresses this limitation. Given a model trained on the source data, we estimate the optimal target anchors in a completely unsupervised manner. The main idea stems from an intuitive observation: by varying the anchor sizes for the target domain, we inevitably introduce noise or even remove valuable object cues. The latent object representation, perturbed by the anchor size, is closest to the learned source features only under the optimal target anchors. We leverage this observation for anchor size optimization. Our experimental results show that, without any retraining, we achieve competitive results even compared to state-of-the-art weakly-supervised size adaptation approaches. In addition, our anchor calibration can be combined with such existing methods, making them completely unsupervised. © 2022, CC BY.

关键词： Anchors

Enforcing connectivity of 3D linear structures using their 2D projections

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Oner, Doruk Osman, Hussein Kozinski, Mateusz Fua, Pascal Computer Vision Laboratory École Polytechnique Fédérale de Lausanne Switzerland Institute of Computer Graphics and Vision Graz University of Technology Austria

Many biological and medical tasks require the delineation of 3D curvilinear structures such as blood vessels and neurites from image volumes. This is typically done using neural networks trained by minimizing voxel-wise loss functions that do not capture the topological properties of these structures. As a result, the connectivity of the recovered structures is often wrong, which lessens their usefulness. In this paper, we propose to improve the 3D connectivity of our results by minimizing a sum of topology-aware losses on their 2D projections. This suffices to increase the accuracy and to reduce the annotation effort required to provide the required annotated training data. Code is available at https://***/doruk-oner/ConnectivityOnProjections. Copyright © 2022, The Authors. All rights reserved.

关键词： Topology

IB-MVS: An Iterative Algorithm for Deep Multi-View Stereo based on Binary Decisions 32

学校读者我要写书评

暂无评论

IB-MVS: An Iterative Algorithm for Deep Multi-View Stereo ba...

32nd British Machine Vision Conference, BMVC 2021

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. © 2021. The copyright of this document resides with its authors.

关键词： Pixels

UAV-based Digital Field Phenotyping for Crop Nitrogen Estimation using RGB Imagery

学校读者我要写书评

暂无评论

UAV-based Digital Field Phenotyping for Crop Nitrogen Estima...

2023 IEEE IAS Global Conference on Emerging Technologies, GlobConET 2023

作者： Patil, Suchitra M Choudhary, Sunita Kholová, Jana Anbazhagan, Krithika Parnandi, Yogesh Gattu, Priyanka Mallayee, Srikanth Ksvv, Prasad Kumar, V Padma Rajalakshmi, P. Chandramouli, Magesh Adinarayana, J. Indian Institute of Technology Bombay Powai Centre of Studies in Resources Engineering Maharashtra Mumbai400076 India Crop Physiology and Modeling Telangana Patancheru502 324 India Czech University of Life Sciences Prague Department of Information Technologies Prague165 00 Czech Republic Indian Institute of Technology Kandi Telangana Hyderabad502285 India International Livestock Research Institute Telangana Patancheru502324 India Purdue University NW Computer Graphics Technology HammondIN46323 United States

ISBN: (纸本)9798350331790

Nitrogen (N) is one of the essential nutrients required for healthy crop growth. Field phenotyping for nitrogen stress symptoms is laborious and time-consuming, that way, it is a major bottleneck in nutrition-inclusive agricultural research. Recent advancements in sensors and image processing facilitate color-based quantification of crop greenness from high-resolution RGB images. In this paper, we present unmanned aerial vehicle (UAV)-based digital field phenotyping for the estimation of crop nitrogen content. For this, we conducted a field experiment during the post-rainy season of 2021 at International Crops Research Institute for Semi-Arid Tropics (ICRISAT), Hyderabad, India with long-stature cereal model crop, sorghum (Sorghum bicolor L.) cultivated under three different regimes varying in moisture and soil nitrogen content. A high-resolution RGB sensor (XenmuseX5S) mounted on DJI Matric 210 quadcopter was used for capturing the spatiotemporal imagery. Five different RGB spectrum vegetation indices indicating crop greenness were correlated with ground truth values of crop N content using simple linear regression and stepwise backward regression. With a prediction potential of R2=0.65 and MAE=0.27 for an independent dataset, we present a stepwise backward linear regression model as a promising approach for real-time estimation of the N status of sorghum crop. © 2023 IEEE.

关键词： Crops

FC-DCNN: A densely connected neural network for stereo estimation

学校读者我要写书评

暂无评论

FC-DCNN: A densely connected neural network for stereo estim...

International Conference on Pattern Recognition

作者： Dominik Hirner Friedrich Fraundorfer Institute for Computer Graphics and Vision Graz University of Technology Austria

We propose a novel lightweight network for stereo estimation. Our network consists of a fully-convolutional densely connected neural network (FC-DCNN) that computes matching costs between rectified image pairs. Our FC-DCNN method learns expressive features and performs some simple but effective postprocessing steps. The densely connected layer structure connects the output of each layer to the input of each subsequent layer. This network structure and the fact that we do not use any fully-connected layers or 3D convolutions leads to a very lightweight network. The output of this network is used in order to calculate matching costs and create a cost-volume. Instead of using time and memory-inefficient cost-aggregation methods such as semiglobal matching or conditional random fields in order to improve the result, we rely on filtering techniques, namely median filter and guided filter. By computing a left-right consistency check we get rid of inconsistent values. Afterwards we use a watershed foreground-background segmentation on the disparity image with removed inconsistencies. This mask is then used to refine the final prediction. We show that our method works well for both challenging indoor and outdoor scenes by evaluating it on the Middlebury, KITTI and ETH3D benchmarks respectively. Our full framework is available at https://***/thedodo/FC-DCNN.

关键词： Matched filters Image segmentation Three-dimensional displays Filtering Neural networks Estimation Benchmark testing

Gaussian Process Emulators for Few-Shot Segmentation in Cardiac MRI 15th

学校读者我要写书评

暂无评论

Gaussian Process Emulators for Few-Shot Segmentation in Ca...

15th International Workshop on Statistical Atlases and Computational Models of the Heart, STACOM 2024, Held in Conjunction with MICCAI 2024

作者： Viti, Bruno Thaler, Franz Kapper, Kathrin Lisa Urschler, Martin Holler, Martin Karabelas, Elias Department of Mathematics and Scientific Computing University of Graz Graz Austria Institute for Medical Informatics Statistics and Documentation Medical University of Graz Graz Austria Institute of Computer Graphics and Vision Graz University of Technology Graz Austria Gottfried Schatz Research Center: Medical Physics and Biophysics Medical University of Graz Graz Austria BioTechMed-Graz Graz Austria

ISBN: (纸本)9783031877551

Segmentation of cardiac magnetic resonance images (MRI) is crucial for the analysis and assessment of cardiac function, helping to diagnose and treat various cardiovascular diseases. Most recent techniques rely on deep learning and usually require an extensive amount of labeled data. To overcome this problem, few-shot learning has the capability of reducing data dependency on labeled data. In this work, we introduce a new method that merges few-shot learning with a U-Net architecture and Gaussian Process Emulators (GPEs), enhancing data integration from a support set for improved performance. GPEs are trained to learn the relation between the support images and the corresponding masks in latent space, facilitating the segmentation of unseen query images given only a small labeled support set at inference. We test our model with the M&Ms-2 public dataset to assess its ability to segment the heart in cardiac magnetic resonance imaging from different orientations, and compare it with state-of-the-art unsupervised and few-shot methods. Our architecture shows higher DICE coefficients compared to these methods, especially in the more challenging setups where the size of the support set is considerably small. The code is available on GitLab (https://***/bruno_viti/gpe_4_cardiac_fss). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Gaussian distribution

Quantifying spatial, temporal, angular and spectral structure of effective daylight in perceptually meaningful ways

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yu, Cehao Wijntjes, Maarten Eisemann, Elmar Pont, Sylvia Faculty of Industrial Design Engineering Delft University of Technology Delft Netherlands Computer Graphics and Visualization Group Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology Delft Netherlands

We present a method to capture the 7-dimensional light field structure, and translate it into perceptually-relevant information. Our spectral cubic illumination method quantifies objective correlates of perceptually relevant diffuse and directed light components, including their variations over time, space, in color and direction, and the environment’s response to sky and sunlight. We applied it "in the wild", capturing how light on a sunny day differs between light and shadow, and how light varies over sunny and cloudy days. We discuss the added value of our method for capturing nuanced lighting effects on scene and object appearance, such as chromatic gradients. © 2022, CC BY.

关键词：