检索结果-内蒙古大学图书馆

16th IEEE International symposium on Mixed and Augmented Reality (ISMAR)

作者： Thiel, Kevin K. Jundt, Eduard Klinker, Gudrun Volkswagen AG Wolfsburg Germany Tech Univ Munich Munich Germany

ISBN: (纸本)9780769563275

Object tracking configuration is a tedious task for users due to an overwhelming amount of parameters while solving a multicriterial optimisation problem for a black box. With increasing complexity of the algorithms and fast changing situations due to the upcoming fourth industrial revolution, even experts will find themselves struggling with this topic. For this reason we present a concept for an automated evaluation and configuration of object tracking algorithms used in augmented reality applications. Given an arbitrary but specified use case the best suitable parameter set shall be found. Therefore the concept utilises statistical analysis of the vast parameter space performing multicriterial optimisation regarding a formalisation of the tracking quality within a virtual testbed.

关键词： I.4.8 [image processing and computer Vision]: Scene Analysis Tracking G.1.6 [Numerical Analysis]: Optimization Global Optimization I.3.7 [computer graphics]: Three-Dimensional graphics and Realism Virtual Reality

来源：评论

学校读者我要写书评

暂无评论

processing-in-Memory Enabled graphics Processors for 3D Rendering 23

Processing-in-Memory Enabled Graphics Processors for 3D Rend...

引用

23rd IEEE International symposium on High Performance computer Architecture (HPCA)

作者： Xie, Chenhao Song, Shuaiwen Leon Wang, Jing Zhang, Weigong Fu, Xin Univ Houston ECE Dept Houston TX 77004 USA PNNL HPC Grp Richland WA USA Capital Normal Univ Coll Informat Engn Beijing Adv Innovat Ctr Imaging Technol Beijing Peoples R China

ISBN: (纸本)9781509049851

The performance of 3D rendering of graphics processing Unit that converts 3D vector stream into 2D frame with 3D image effects significantly impacts users gaming experience on modern computer systems. Due to its high texture throughput requirement, main memory bandwidth becomes a critical obstacle for improving the overall rendering performance. 3D-stacked memory systems such as Hybrid Memory Cube provide opportunities to significantly overcome the memory wall by directly connecting logic controllers to DRAM dies. Although recent works have shown promising improvement in performance by utilizing HMC to accelerate special-purpose applications, a critical challenge of how to effectively leverage its high internal bandwidth and computing capability in GPU for 3D rendering remains unresolved. Based on the observation that texel fetches greatly impact off-chip memory traffic, we propose two architectural designs to enable processing-In-Memory based GPU for efficient 3D rendering. Additionally, we employ camera angles of pixels to control the performance-quality tradeoff of 3D rendering. Extensive evaluation across several real-world games demonstrates that our design can significantly improve the performance of texture filtering and 3D rendering by an average of 3.97X (up to 6.4X) and 43% (up to 65%) respectively, over the baseline GPU. Meanwhile, our design provides considerable memory traffic and energy reduction without sacrificing rendering quality.

关键词： GPU processing-In-Memory 3D-Stacked Memory Approximate Computing 3D Rendering

来源：评论

学校读者我要写书评

暂无评论

NASA Technical Reports Server (Ntrs) 19900007126: the 2Nd symposium on the Frontiers of Massively Parallel Computations

引用

2017年

NASA Technical Reports Server (Ntrs) 19900007126: the 2Nd symposium on the Frontiers of Massively Parallel Computations by NASA Technical Reports Server (Ntrs); published by

关键词： (ntrs) 19900007126: charged particles computations computer graphics conferences frontiers image processing massively mills, ronnie [editor] mimd (computers) nasa technical reports server (ntrs) parallel processing (computers) programming languages simd (computers) supercomputers

来源：评论

学校读者我要写书评

暂无评论

Group Immersive Education with Digital Fulldome Planetariums 19

Group Immersive Education with Digital Fulldome Planetariums

引用

19th IEEE Virtual Reality Conference (VR)

作者： Yu, Ka Chun Sahami, Kamran Sahami, Victoria Sessions, Larry Denn, Grant Denver Museum Nat & Sci Denver CO USA Metropolitan State Univ Denver Denver CO USA

ISBN: (纸本)9781509066476

Although fulldome video digital theaters evolved from traditional planetariums, they are more akin to virtual reality (VR) theaters that create large-scale, group immersive experiences. In order to help understand how immersion and wide fields-of-view (FOV) impact learning, we studied the use of visualizations on topics that do and do not require spatial understanding in astronomy classes. We find a significant difference between students who viewed visualizations in the dome versus those that saw non-immersive content in their classrooms, with the former showing the greatest retention. Our results suggest that immersive visuals help free up cognitive resources that can be used to build mental models requiring spatial understanding, and the physical display size combined with the wide FOV may result in greater attention. Although fulldome is a complementary medium to traditional VR, our results have implications for future head-mounted displays.

关键词： 1.3.7 [computer graphics]: Three-Dimensional graphics and Realism-Virtual reality 1.4.0 [image processing and computer Vision]: General-image displays

来源：评论

学校读者我要写书评

暂无评论

Real-Time Face Relighting via Adaptive Normal Mapping 17

Real-Time Face Relighting via Adaptive Normal Mapping

引用

ACM-SIGGRAPH Asia computer Animation Festival Conference / SIGGRAPH Asia Courses Conference / SIGGRAPH Asia Emerging Technologies Conference / symposium on Mobile graphics and Interactive Applications

作者： Choudhary, Pradeep Mannar, Sumedh Kumar, Amit Son, Minjung Samsung Res Inst India Bangalore Karnataka India Samsung Adv Inst Technol Suwon South Korea

ISBN: (纸本)9781450354103

In this paper, we present a real-time face relighting method for enhancing widely used face photographs such as 'selfies'. In contrast with previous relighting methods that estimate and replace the real lighting condition with the intended one, our technique focuses on the aesthetic aspects of relighting with real-time interaction. We estimate the head pose based on the facial features and apply adaptive face normal mapping by transforming and interpolating pre-defined face normal maps. The lighting effect is calculated using adaptively mapped normals and overlaid on the original photograph. The experimental results show that the proposed method generates aesthetic face photographs without a computational overhead, enabling real-time face relighting for mobile applications.

关键词： image and video processing image and video enhancement real-time mobile application image-based relighting face relighting

来源：评论

学校读者我要写书评

暂无评论

A GPU-based convolutional neural network approach for image classification

A GPU-based convolutional neural network approach for image ...

引用

2017 International Artificial Intelligence and Data processing symposium, IDAP 2017

作者： Cengil, Emine Çinar, Ahmet Güler, Zafer Departure of Computer Engineering Firat University Elazig Turkey

ISBN: (纸本)9781538618806

Deep learning obtains successful results in solving many machine learning problems. In this study, image classification process is performed by using Convolutional Neural Network (CNN) which is the most used architecture of deep learning. image classification is used in a lot of basic field like medicine, education and security. Conditions that correct classification has vital importance may be especially in medicine field. Therefore, improved methods are needed in this issue. Although several algorithms for image classification have been developed over the years, they have not been used with the discovery of Convolutional Neural Networks. Convolutional Neural Networks provide better results than existing methods in the literature due to advantages such as processing by extracting hidden features, allowing parallel processing thanks to parallel structure, and real time operation. Furthermore, we use Convolutional Neural Networks in the proposed method. In this study, the image classification process is performed by using like a LeNet network model. The caffe library, which is often used for deep learning, is utilized. Our method is trained and tested with images of cats and dogs taken from the kaggle dataset. 10.000 tagged data is used for training and 5.000 unlabeled data is used for testing. Owing to Convolutional Neural Networks allow parallel processing, GPU technology has been used. In our method is used GPU technology and classification is evaluated with acceptable accuracy rate and speed performance. © 2017 IEEE.

关键词： graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Bayesian Collaborative Denoising for Monte Carlo Rendering

引用

computer graphics FORUM 2017年第4期36卷 137-153页

作者： Boughida, Malik Boubekeur, Tamy Paris Saclay Univ LTCI Telecom ParisTech Paris France

The stochastic nature of Monte Carlo rendering algorithms inherently produces noisy images. Essentially, three approaches have been developed to solve this issue: improving the ray-tracing strategies to reduce pixel variance, providing adaptive sampling by increasing the number of rays in regions needing so, and filtering the noisy image as a post-process. Although the algorithms from the latter category introduce bias, they remain highly attractive as they quickly improve the visual quality of the images, are compatible with all sorts of rendering effects, have a low computational cost and, for some of them, avoid deep modifications of the rendering engine. In this paper, we build upon recent advances in both non-local and collaborative filtering methods to propose a new efficient denoising operator for Monte Carlo rendering. Starting from the local statistics which emanate from the pixels sample distribution, we enrich the image with local covariance measures and introduce a nonlocal bayesian filter which is specifically designed to address the noise stemming from Monte Carlo rendering. The resulting algorithm only requires the rendering engine to provide for each pixel a histogram and a covariance matrix of its color samples. Compared to state-of-the-art sample-based methods, we obtain improved denoising results, especially in dark areas, with a large increase in speed and more robustness with respect to the main parameter of the algorithm. We provide a detailed mathematical exposition of our bayesian approach, discuss extensions to multiscale execution, adaptive sampling and animated scenes, and experimentally validate it on a collection of scenes.

关键词： CCS Concepts •Computing methodologies Ray tracing image processing

来源：评论

学校读者我要写书评

暂无评论

A 2D/3D environment perception approach applied to sensor-based navigation of automated driving systems 14

A 2D/3D environment perception approach applied to sensor-ba...

引用

14th Latin American Robotics symposium and 5th brazilian symposium on Robotics, LARS-SBR 2017

作者： De Lima, Danilo Alves Victorino, Alessandro Corrêa De Miranda Neto, Arthur Terrestrial Mobility Lab. Federal University of Lavras Brazil Heudiasyc UMR CNRS 7253 University of Technology of Compiègne France

ISBN: (纸本)9781538609569

The perception of the environment is one the most important tasks related to automated driving systems navigation. Typically, the robot detects the road surface and obstacles to perform the local navigation control safely. However, working with image data brings several problems related to environment elements like shadows, light reflection, low-variance textures, etc., which could compromise the efficiency of the final detection. We dealt with these problems in our previous work, where a 2D/3D image processing approach was proposed to detect the road surface. Although presenting good results, the final solution was time-consuming for a real application in autonomous navigation. Here, we propose some modifications in the 2D/3D structure to a better time efficient technique, allowing the road surface and obstacles detection. The final solution was tested successfully in a sensor-based navigation approach. © 2017 IEEE.

关键词： image segmentation

来源：评论

学校读者我要写书评

暂无评论

EXTRACTION OF BUILDINGS IN VHR SAR imageS USING FULLY CONVOLUTION NEURAL NETWORKS

EXTRACTION OF BUILDINGS IN VHR SAR IMAGES USING FULLY CONVOL...

引用

IEEE International Geoscience and Remote Sensing symposium

作者： Muhammad Shahzad Michael Maurer Friedrich Fraundorfer Yuanyuan Wang Xiao Xiang Zhu Signal Processing in Earth Observation (SiPEO) Technische Universität München (TUM) Munich Germany Technical University of Graz (TU Graz) Institute for Computer Graphics and Vision (ICG) Graz Austria

Modern spaceborne synthetic aperture radar (SAR) sensors, such as TerraSAR-X/TanDEM-X and COSMO-SkyMed, can deliver very high resolution (VHR) data beyond the inherent spatial scales (on the order of 1m) of buildings, constituting invaluable data source for large-scale urban mapping. processing this VHR data with advanced interferometric techniques, such as SAR tomography (TomoSAR), enables the generation of 3-D (or even 4-D) TomoSAR point clouds from space. In this paper, we present a novel and generic workflow that exploits these TomoSAR point clouds in a way that is capable to automatically produce benchmark annotated (buildings/non-buildings) SAR datasets. These annotated datasets (building masks) have been utilized to construct and train the state-of-the-art deep Fully Convolution Neural Networks with an additional Conditional Random Field represented as a Recurrent Neural Network to detect building regions in a single VHR SAR image. The results of building detection are illustrated and validated over TerraSAR-X VHR spotlight SAR image covering approximately 39 km~2 - almost the whole city of Berlin - with mean pixel accuracies of around 93.84%.

关键词： Synthetic aperture radar Buildings Three-dimensional displays Training Urban areas Tomography Network architecture

来源：评论

学校读者我要写书评

暂无评论

A Method of View-Dependent Stereoscopic Projection on Curved Screen

A Method of View-Dependent Stereoscopic Projection on Curved...

引用

IEEE Annual International symposium Virtual Reality

作者： Juan Liu Hanchao Li Lu Zhao Siwei Zhao Guowen Qi Yulong Bian Xiangxu Meng Chenglei Yang Shandong University Jinan China Zhejiang University Hangzhou China

ISBN: (纸本)9781538633663

In this paper, we present a method of view-dependent stereoscopic projection on curved screen. It allows the user to walk around with the correct perspective view of the virtual scene consistent with his/her location. To solve the problem of distortion and drift of virtual objects when projecting the view-dependent scene images on curved screen, we operate a dynamic parallax adjustment of the stereoscopic images according to viewpoints. User evaluation shows that our proposed approaches are effective on improving visual experience.

关键词： Stereo image processing Visualization Three-dimensional displays Head Cameras Rendering (computer graphics)

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：