高效KD树并行算法优化
作者:
基金项目:

上海市科委科技攻关项目(13DZ1108800);国家高技术研究发展计划(863)(2012AA010901);国家自然科学基金(61370081)


Parallelization Optimization of KD-Tree Building Algorithm
Author:
  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献 [23]
  • |
  • 相似文献 [20]
  • | | |
  • 文章评论
    摘要:

    KD树作为一种用于查询高维键值的流行算法, 由于其准确性高、可扩展性强与较快的查询速度而应用于多媒体检索领域, 但缓慢的建树效率已不能很好的满足当前的应用场景. 针对KD树的低效建树过程, 作者探寻并分析了KD树建树现存的并行潜能并提出了一种面向KD树建树过程的多核并行算法—ParK(Parallel KD-Tree). ParK探求了不同的并行模式来充分利用现代硬件中的计算资源, 并在此基础上提出了一种新的内存分配策略来解决并行处理中的数据争用状况. 实验结果表明Park相比于原始串行版本最高能够在16核的服务器上达到21.75倍的加速.

    Abstract:

    In recent years, multimedia data has become one of the major data types transferred and processed on the Internet. K dimensional tree is one of the most popular tree structures for searches involving a multidimensional search key, which is similar to feature points extracted from multimedia data, due to its good accuracy, scalability and fast retrieval speed. However, its slow building speed limits its application area, especially with large dataset. Fortunately, Modern processors provide tremendous computing power by integrating multiple or many cores. In this paper, we explore and analyze the existing potential parallel in KD-Tree building process. Then we present ParK, a customized parallel solution that exploits multi-core CPUs to accelerate KD-Tree building process. ParK exploits different parallel models to fully utilize computation resource in modern hardware and solves data race by presenting a new memory allocation strategy. The final experimental results show ParK achieves about 21.75X speedup compared to original serial version on 16-core server.

    参考文献
    1 Youtube statistics. http://www.youtube.com/t/press statistics.
    2 Philbin J, Chum O, Isard M, et al. Object retrieval with large vocabularies and fast spatial matching. IEEE Conference on Computer Vision and Pattern Recognition(CVPR'07). IEEE. 2007. 1-8.
    3 Mikolajczyk K, Uemura H. Action recognition with motion- appearance vocabulary forest. IEEE Conference on Computer Vision and Pattern Recognition(CVPR 2008). IEEE. 2008. 1-8.
    4 Bentley JL. Multidimensional binary search trees used for associative searching. Communications of the ACM, 1975, 18(9): 509-517.
    5 Mikolajczyk K, Matas J. Improving descriptors for fast tree matching by optimal linear projection. IEEE 11th International Conference on Computer Vision(ICCV 2007). IEEE. 2007. 1-8.
    6 Lowe DG. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, 2004, 60(2): 91-110.
    7 Bay H, Tuytelaars T, Van Gool L. Surf: Speeded up robust features. Computer Vision-ECCV 2006. Springer Berlin Heidelberg, 2006: 404-417.
    8 Vedaldi A, Fulkerson B. VLFeat: An open and portable library of computer vision algorithms. Proc. of the International Conference on Multimedia. ACM. 2010. 1469-1472.
    9 Nister D, Stewenius H. Scalable recognition with a vocabulary tree. IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE. 2006, 2. 2161-2168.
    10 Gionis A, Indyk P, Motwani R. Similarity search in high dimensions via hashing. VLDB, 1999, 99: 518-529.
    11 Datar M, Immorlica N, Indyk P, et al. Locality-sensitive hashing scheme based on p-stable distributions. Proc. of the Twentieth Annual Symposium on Computational Geometry. ACM. 2004. 253-262.
    12 Fischler MA, Bolles RC. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Communications of the ACM, 1981, 24(6): 381-395.
    13 McNames J. A fast nearest-neighbor algorithm based on a principal axis search tree. IEEE Trans. on, Pattern Analysis and Machine Intelligence, 2001, 23(9): 964-976.
    14 Sproull RF. Refinements to nearest-neighbor searching ink-dimensional trees. Algorithmica, 1991, 6(1-6): 579-589.
    15 Zhang N. Computing parallel speeded-up robust features (P-SURF) via POSIX threads. Emerging Intelligent Computing Technology and Applications. Springer Berlin Heidelberg, 2009: 287-296.
    16 Chen P, Yang D, Zhang W, et al. Adaptive pipeline parallelism for image feature extraction algorithms. 41st International Conference on Parallel Processing(ICPP). 2012. IEEE. 299-308.
    17 Warn S, Emeneker W, Cothren J, et al. Accelerating SIFT on parallel architectures. CLUSTER. 2009. 1-4.
    18 Heymann S, Müller K, Smolic A, et al. SIFT implementation and optimization for general-purpose GPU. 2007.
    19 Cornelis N, Van Gool L. Fast scale invariant feature detection and matching on programmable graphics hardware. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops(CVPRW'08). IEEE. 2008. 1-8.
    20 Terriberry TB, French LM, Helmsen J. GPU accelerating speeded-up robust features. Proc. Int. Symp. on 3D Data Processing, Visualization and Transmission (3DPVT). 2008. 355-362.
    21 Havran V, Bittner J. On improving KD-trees for ray shooting. Proc. of Wscg Conference. 2002. 209-217.
    22 Di Fatta G, Pettinger D. Dynamic load balancing in parallel KD-Tree K-Means. 2010 IEEE 10th International Conference on Computer and Information Technology(CIT). IEEE. 2010. 2478-2485.
    23 Zhou P, Meng X. SAH based KD tree construction on hybrid architecture. 2011 Workshop on Digital Media and Digital Content Management (DMDCM). IEEE. 2011. 185-189.
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

李天驹,张铮,张为华.高效KD树并行算法优化.计算机系统应用,2015,24(8):1-9

复制
分享
文章指标
  • 点击次数:2342
  • 下载次数: 4376
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 收稿日期:2014-12-03
  • 最后修改日期:2015-01-12
  • 在线发布日期: 2015-09-03
文章二维码
您是第12693705位访问者
版权所有:中国科学院软件研究所 京ICP备05046678号-3
地址:北京海淀区中关村南四街4号 中科院软件园区 7号楼305房间,邮政编码:100190
电话:010-62661041 传真: Email:csa (a) iscas.ac.cn
技术支持:北京勤云科技发展有限公司

京公网安备 11040202500063号