检索结果-内蒙古大学图书馆

IEEE International Workshop/Symposium on Workload Characterization

作者： Zhibin Yu Hai Jin Nilanjan Goswami Tao Li Lizy K. John Service Computing Technologies and System Lab/Cluster and Grid Computing Lab Huazhong University of Science and Technology Wuhan China Intelligent Design of Efficient Architecture Lab University of Florida Gainesville FL USA Department of Electrical and Computer Engineering University of Texas Austin Austin TX USA

CUDA has become a very popular programming paradigm in parallel computing area. However, very little work has been done for characterizing CUDA kernels. In this work, we measure the thread level performance, collect the basic block level characteristics, and glean the instruction level properties for about 35 programs from CUDA SDK, Parboil, and Rodinia benchmark suites. In addition, we define basic block vectors, synchronization vectors and thread similarity matrix to capture the characteristics of CUDA programs efficiently. We find that CUDA programs have some unique characteristics at each level compared to sequential programs.

关键词： Graphics processing unit Vectors Kernel Instruction sets Programming USA Councils Educational institutions

来源：评论

学校读者我要写书评

暂无评论

Creating the perception-based LADDER sketch recognition language 10

Creating the perception-based LADDER sketch recognition lang...

引用

Proceedings of the 8th ACM Conference on designing Interactive Systems

作者： Hammond, Tracy Davis, Randall Sketch Recognition Lab. Department of Computer Science and Engineering Texas A and M University United States Design Rationale Group Computer Science and Artificial Intelligence Lab. Massachusetts Institute of Technology United States

ISBN: (纸本)9781450301039

Sketch recognition is automated understanding of hand-drawn diagrams. Current sketch recognition systems exist for only a handful of domains, which contain on the order of 10 - 20 shapes. Our goal was to create a generalized method for recognition that could work for many domains, increasing the number of shapes that could be recognized in real-time, while maintaining a high accuracy. In an effort to effectively recognize shapes while allowing drawing freedom (both drawing-style freedom and perceptually-valid variations), we created the shape description language modeled after the way people naturally describe shapes to 1) create an intuitive and easy to understand description, providing transparency to the underlying recognition process, and 2) to improve recognition by providing recognition flexibility (drawing freedom) that is aligned with how humans perceive shapes. This paper describes the results of a study performed to see how users naturally describe shapes. A sample of 35 subjects described or drew approximately 16 shapes each. Results show a common vocabulary related to Gestalt grouping and singularities. Results also show that perception, similarity, and context play an important role in how people describe shapes. This study resulted in a language (LADDER) that allows shape recognizers for any domain to be automatically generated from a single hand-drawn example of each shape. Sketch systems for over 30 different domains have been automatically generated based on this language. The largest domain contained 923 distinct shapes, and achieved a recognition accuracy of 83% (and a top-3 accuracy of 87%) on a corpus of over 11,000 sketches, which recognizes almost two orders of magnitude more shapes than any other existing system. © 2010 ACM.

关键词： Ladders

来源：评论

学校读者我要写书评

暂无评论

Impact of colinearity of sensors selected for location estimation

Impact of colinearity of sensors selected for location estim...

引用

17th International Conference on High Performance Computing, HiPC 2010

作者： Sadaphal, Vaishali P. Jain, Bijendra N. Systems Research Lab. Tata Research Development and Design Centre Pune India Dept. of Computer Science and Engineering Indian Institute of Technology Delhi India

ISBN: (纸本)9781424485185

We consider estimating the location of a target moving in a 2D plane by combining distance measurements from multiple sensors. Given that available energy in sensors is at a premium, it is desirable that energy be conserved by selecting fewer number of sensors that measure distance and communicate to the central tracker. We propose heuristics on the basis of which a handful of sensors may be selected. In this paper, we present one of the heuristic. The particular heuristic discussed at length is based on colinearity of sensors. In this paper, we provide a formal and theoretical basis for arriving at the heuristic and evaluate its utility. ©2010 IEEE.

关键词： distance measurement distance measurement location estimation moving target sensor colinearity sensor energy sensor fusion target tracking

来源：评论

学校读者我要写书评

暂无评论

Optimal rectangle packing

引用

Annals of Operations Research 2010年第1期179卷 261-295页

作者： Korf, Richard E. Moffitt, Michael D. Pollack, Martha E. Computer Science Department University of California Los Angeles CA 90095 United States Design Productivity Group IBM Austin Research Lab Austin TX 78758 United States Department of Electrical Engineering and Computer Science University of Michigan Ann Arbor MI 48109 United States

We consider the NP-complete problem of finding an enclosing rectangle of minimum area that will contain a given a set of rectangles. We present two different constraint-satisfaction formulations of this problem. The first searches a space of absolute placements of rectangles in the enclosing rectangle, while the other searches a space of relative placements between pairs of rectangles. Both approaches dramatically outperform previous approaches to optimal rectangle packing. For problems where the rectangle dimensions have low precision, such as small integers, absolute placement is generally more efficient, whereas for rectangles with high-precision dimensions, relative placement will be more effective. In two sets of experiments, we find both the smallest rectangles and squares that can contain the set of squares of size 1 × 1, 2 × 2, ..., N × N, for N up to 27. In addition, we solve an open problem dating to 1966, concerning packing the set of consecutive squares up to 24×24 in a square of size 70×70. Finally, we find the smallest enclosing rectangles that can contain a set of unoriented rectangles of size 1 × 2, 2 × 3, 3 × 4, ..., N × (N+1), for N up to 25. © 2008 Springer science+Business Media, LLC.

关键词： Constraint satisfaction Rectangle packing Search

来源：评论

学校读者我要写书评

暂无评论

Thermal aware test scheduling for stacked multi-chip-modules

Thermal aware test scheduling for stacked multi-chip-modules

引用

East-West design and Test Symposium

作者： Vinay, N.S. Rawaty, Indira Larssonz, Erik Gaurx, M.S. Singh, Virendra Cisco Systems Bangalore India Electrical Enggineering Dept. Engineering College Ajmer India Dept. of Computer Science Linkoping University Linkoping Sweden Computer Engineering Dept. Malaviya National Institute of Technology Jaipur India Computer Design and Test Lab. Indian Institute of Science Bangalore India

ISBN: (纸本)9781424495566

In an attempt to increase the area utilization of multi-chip packages, manufacturers have started looking at 3D packaging. The 3D structures have either dies or chips stacked one above the other. These 3D structures have low thermal capabilities, and hence the thermal issues get aggravated during testing. In this paper, thermal aware test scheduling techniques for stacked structures is proposed. Care is taken to schedule the stacked partitions appropriately to prevent local heating and achieve uniform vertical temperature spread. Concurrent testing of dies is considered for reduced test time and existing test architecture is employed. The widely used global peak power model gives a pessimistic estimate of the power profile. A new partition scheme is proposed for the power profile based on power variations and true power. This scheme of power profile partitioning which is a trade-off between a cycle accurate model and global peak model offers both flexibility during scheduling and also reduces total false power. The power profile partitions are scheduled through algorithms based on heuristics. System model of the 3D stack is obtained from the linear RC-models after considering the various factors that affect the heat flow and die temperatures. Simulation results show that the proposed technique achieves uniform temperature spread across the 3D structure. © 2010 IEEE.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

design contest overview: Combined architecture for network stream categorization and intrusion detection (CANSCID) 10

Design contest overview: Combined architecture for network s...

引用

Proceedings of the Eighth ACM/IEEE International Conference on Formal Methods and Models for Codesign

作者： Pellauer, Michael Agarwal, Abhinav Khan, Asif Ng, Man Cheuk Vijayaraghavan, Muralidaran Brewery, Forrest Emer, Joel Massachusetts Institute of Technology Computation Structures Group Computer Science and AI Lab. Cambridge MA United States University of California Santa Barbara High Level Design Group Electrical and Computer Engineering Santa Barbara CA United States Intel Corporation VSSAD Group Hudson MA United States

ISBN: (纸本)9781424478859

This year we received 8 submissions for our Deep Packet Inspection problem. 6 submissions used FPGAs, and 2 used GP-GPUs. The organizers find it significant that no team submitted a software-only solution that did not use some kind of hardware accelerator an indication that software alone could not meet the required line rate. This year the contest ended in a tie. Congratulations to the joint winners, Team Sasao lab and Team Limenators, each having implemented 140 patterns while maintaining line rate. Additionally, Team Sasao lab was the only team to use an NFA approach rather than DFAs for matching the regular expressions. Full results are given in Table II. The performance of the two winners was verified by the organizers using undisclosed test inputs. The performance of the other teams is self-reported. © 2010 IEEE.

关键词： Intrusion detection

来源：评论

学校读者我要写书评

暂无评论

CRISTAL: A Collaborative Home Media and Device Controller Based on a Multi-touch Display 09

CRISTAL: A Collaborative Home Media and Device Controller Ba...

引用

4th ACM International Conference on Interactive Tabletops and Surfaces, ITS 2009

作者： Seifried, Thomas Haller, Michael Scott, Stacey D. Perteneder, Florian Rendl, Christian Sakamoto, Daisuke Inami, Masahiko Media Interaction Lab Upper Austria University of Applied Sciences Austria Department of Systems Design Engineering University of Waterloo Canada Department of Computer Science University of Tokyo Japan School of Media Design Keio University Japan

ISBN: (纸本)9781605587332

While most homes are inherently social places, existing devices designed to control consumer electronics typically only support single user interaction. Further, as the number of consumer electronics in modern homes increases, people are often forced to switch between many controllers to interact with these devices. To simplify interaction with these devices and to enable more collaborative forms of device control, we propose an integrated remote control system, called CRISTAL (Control of Remotely Interfaced Systems using Touch-based Actions in Living spaces). CRISTAL enables people to control a wide variety of digital devices from a centralized, interactive tabletop system that provides an intuitive, gesture-based interface that enables multiple users to control home media devices through a virtually augmented video image of the surrounding environment. A preliminary user study of the CRISTAL system is presented, along with a discussion of future research directions. © 2009 ACM.

关键词： Digital devices

来源：评论

学校读者我要写书评

暂无评论

Introducing new associate editors

引用

Journal of Computing and Information science in engineering 2009年第3期9卷 1-2页

作者： Saitou, Kazuhiro Grosse, Ian R. Gao, Shuming McMains, Sara Qian, Xiaping Goel, Ashok Geiger, Chris NSF Center for e-Design Mechanical and Industrial Engineering Univsersity of Massachusetts Amherst United States State Key Lab of CAD and CG Zhejiang University Hangzhou China Mechanical Engineering UC Berkeley United States Illinois Institute of Technology Chicago United States Computer and Cognitive Science School of Interactive Computing Georgia Institute of Technology United States University of Applied Sciences Düsseldorf Germany

来源：评论

学校读者我要写书评

暂无评论

Fast intra prediction mode decision algorithm for 4×4 blocks in H.264/MPEG-4 AVC

引用

Tamkang Journal of science and engineering 2008年第1期11卷 49-54页

作者： Chiang, Jen-Shiun Hsia, Chih-Hsien Teng, Tsai-Yuan Multimedia IC Design Lab. Department of Electrical Engineering Tamkang University Tamsui 251 Taiwan Department of Computer Science and Information Engineering Tamkang University Tamsui 251 Taiwan

The H.264/MPEG-4 (Advanced Video Coding) AVC video coding includes a multidirectional spatial prediction method to reduce spatial redundancy by using neighboring samples as a prediction for the samples in a block of data to be encoded. The main goal of the intra prediction is to achieve better compression efficiency. For finding the optimal prediction modes, H.264/MPEG-4 AVC adopts the full search algorithm in the standard. The high computation complexity of the full search algorithm makes the encoding of H.264/MPEG-4 AVC to be difficult to meet real time applications. In this work, we proposed a Fast Intra Prediction Mode Decision Algorithm (FIPMDA) for 4 × 4 intra blocks, which is based on the partially sampling prediction and symmetry of the adjacent angle modes, to reduce the computational time and low complexity of the 4 × 4 block intra-prediction. Experimental results show that the proposed method can reduce the encoder time of the intra prediction by more than 60% while maintaining similar video bit rate and quality.

关键词： Image coding

来源：评论

学校读者我要写书评

暂无评论

Advances in tangible interaction and ubiquitous virtual reality

Advances in tangible interaction and ubiquitous virtual real...

引用

作者： Hong, Dongpyo Höllerer, Tobias Haller, Michael Takemura, Haruo Cheok, Adrian David Kim, Gerard Jounghyun Billinghurst, Mark Woo, Woontack Hornecker, Eva Jacob, Robert J.K. Hummels, Caroline Ullmer, Brygg Schmidt, Albrecht Van Den Hoven, Elise Mazalek, Ali Gwangju Institute of Science and Technology Department of Computer Science University of California Santa Barbara Upper Austria University of Applied Sciences Department of Digital Media Infomedia Education Division Osaka University Cybermedia Center National University of Singapore Mixed Reality Lab. Keio University Department of Computer Science and Engineering Korea University Gwangju Institute of Science and Technology Open University Pervasive Interaction Lab. Department of Computer Science Tufts University Eindhoven University of Technology Department of Industrial Design Department of Computer Science Louisiana State University University of Duisburg Pervasive Computing and User Interface Engineering Group Essen Digital Media Program Georgia Institute of Technology

The first article reports on context-sensitive augmented-reality research presented at the 2007 International Symposium on Ubiquitous Virtual Reality. This student-organized event explored the use of contextual information, design principles, and effective user evaluation for developing AR applications for ubiquitous computing environments. The second article reports on The International Conference on Tangible and Embedded Interaction, the first conference series worldwide to focus on tangible and embedded interaction. The conference is interdisciplinary, covering the arts, hardware design, software toolkits for prototyping, and user studies and theory development. © 2006 IEEE.

关键词： Virtual reality

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：