计算机科学 ›› 2016, Vol. 43 ›› Issue (9): 39-46, 51.doi: 10.11896/j.issn.1002-137X.2016.09.007

• 2015 年第三届CCF 大数据学术会议 • 上一篇    下一篇

FP-CNNH:一种基于深度卷积神经网络的快速图像哈希算法

刘冶,潘炎,夏榕楷,刘荻,印鉴   

  1. 中山大学数据科学与计算机学院 广州510000;中山大学信息科学与技术学院 广州510000,中山大学数据科学与计算机学院 广州510000;中山大学软件学院 广州510000,中山大学信息科学与技术学院 广州510000,中山大学数据科学与计算机学院 广州510000,中山大学数据科学与计算机学院 广州510000;中山大学信息科学与技术学院 广州510000
  • 出版日期:2018-12-01 发布日期:2018-12-01
  • 基金资助:
    本文受国家自然科学基金(61033010,5,61370021,3,U1401256),广东省自然科学基金(S2011020001182,S2012010009311,S2013010011905),广东省科技计划项目(2011B040200007,2A010701013),火烈鸟网络(广州)股份有限公司-中山大学广东省大数据分析与处理重点实验室校企产学研合作研究经费资助

FP-CNNH:A Fast Image Hashing Algorithm Based on Deep Convolutional Neural Network

LIU Ye, PAN Yan, XIA Rong-kai, LIU Di and YIN Jian   

  • Online:2018-12-01 Published:2018-12-01

摘要: 在大数据时代,图像检索技术在大规模数据上的应用是一个热门的研究领域。近年来,大规模图像检索系统中, 图像哈希算法 由于具备提高图像的检索效率同时减少储存空间的优点而受到广泛的关注。现有的有监督学习哈希算法存在一些问题,主流的有监督的哈希算法需要通过图像特征提取器获取人为构造的图像特征表示,这种做法带来的图像特征损失影响了哈希算法的效果,也不能较好地处理图像数据集中语义的相似性问题。随着深度学习在大规模数据上研究的兴起,一些相关研究尝试通过深度神经网络进行有监督的哈希函数学习,提升了哈希函数的效果,但这类方法需要针对数据集人为设计复杂的深度神经网络,增大了哈希函数设计的难度,而且深度神经网络的训练需要较多的数据和较长的时间,这些问题影响了基于深度学习的哈希算法在大规模数据集上的应用。针对这些问题,提出了一种基于深度卷积神经网络的快速图像哈希算法,该算法通过设计优化问题的求解方法以及使用预训练的大规模深度神经网络,提高了哈希算法的效果,同时明显地缩短了复杂神经网络的训练时间。根据在不同图像数据集上的实验结果分析可知, 与现有的基准算法相比,提出的算法在哈希函数训练效果和训练时间上都具有较大的提高。

关键词: 深度学习,图像检索,图像哈希,神经网络,优化算法

Abstract: In the big data era,application research on image retrieval technology for large-scale data is a hot field.Recent years,image hash algorithm has attracted much attention in large-scale image retrieval system in order to improve the retrieval efficiency and reduce the storage space.However,there are some issues with existing supervised hash code learning algorithms.Most of the supervised hash algorithms need to use image feature extractor for obtaining hand-crafted image features,which influence the effect of hash code training with image features,at the same time these methods cannot deal well with semantic similarity for image data set.With the development of deep learning research on the large-scale data,some recent related work try to deploy deep neural network to learn hash function and make the hash code training effect increased.But such kind of methods require carefully designed complex neural network structure thus increase the difficulty of the hash function design and cost more time on neural network training with large data set.These problems limit the range of the hash algorithm application with the deep learning architecture for large data sets.To solve the problems mentioned above,this paper proposed a fast image hashing algorithm based on deep convolution neural network.The proposed algorithm consists of an optimization approach for constructing the hash code of the training data set and a pre-trained large deep neural network for learning to improve the effect of hash algorithm,shor-tening the training time of complex neural network.According to the analysis of experimental results on different image data sets,both the effectiveness of hash function and the efficiency of training time of the proposed algorithm have better performance compared with the existing algorithms.

Key words: Deep learning,Image retrieval,Image hashing,Neural network,Optimization algorithm

[1] Jégou H,Douze M,Schmid C,et al.Aggregating local descriptors into a compact image representation[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2010:3304-3311
[2] Charikar M S.Similarity estimation techniques from rounding algorithms[C]∥Proceedings of the Symposium on Theory of Computing.2002:380-388
[3] Xia R,Pan Y,Lai H,et al.Supervised Hashing for Image Retrieval via Image Representation Learning[C]∥Proceedings of AAAI Conference on Artificial Intelligence (AAAI).2014
[4] Gong Y,Lazebnik S.Iterative quantization:A procrustean approach to learning binary codes[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2011:817-824
[5] Wang Hui-feng,Sun Zheng-xing,Wang Jian.Semantic ImageRetrieval:Review And Research[J].Journal of Computer Research and Development,2002,9(5):513-523(in Chinese) 王惠锋,孙正兴,王箭.语义图像检索研究进展[J].计算机研究与发展,2002,39(5):513-523
[6] Wu Hong,Lu Han-qing,Ma Song-de.A Survey of RelevanceFeedback Techniques in Content-Based Image Retrieval[J].Chinese Journal of Computers,2005,8(12):1969-1979(in Chinese) 吴洪,卢汉清,马颂德.基于内容图像检索中相关反馈技术的回顾[J].计算机学报,2005,28(12):1969-1979
[7] Jegou H,Douze M,Schmid C.Hamming embedding and weak geometric consistency for large scale image search[C]∥Proceedings of the European Conference on Computer Vision.2008:304-317
[8] Gionis A,Indyk P,Motwani R.Similarity search in high dimensions via hashing[C]∥Proceedings of the International Confe-rence on Very Large Data Bases.1999,99:518-529
[9] Kleinberg J M.Two algorithms for nearest-neighbor search in high dimensions[C]∥Proceedings of the Symposium on Theory of Computing.New York:ACM Press,1997:599-608
[10] Kulis B,Grauman K.Kernelized locality-sensitive hashing[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(6):1092-1104
[11] Weiss Y,Torralba A,Fergus R.Spectral Hashing[J].Procee-dings of the Advances in Neural Information Processing Systems.2008,283(3):1753-1756
[12] Arya S,Mount D M,Netanyahu N S,et al.An optimal algorithm for approximate nearest neighbor searching fixed dimensions[J].Journal of the ACM,1998,45(6):891-923
[13] Lai H,Pan Y,Liu Y,et al.Simultaneous Feature Learning and Hash Coding with Deep Neural Networks[C]∥Proceedings of IEEE International Conference on Pattern Recognition and Computer Vision (CVPR).2015
[14] Clarkson K L.An algorithm for approximate closest-point queries[C]∥Proceedings of the Symposium on Computational Geo-metry.1994:160-164
[15] Wang J,Kumar S,Chang S F.Semi-supervised hashing for scalable image retrieval[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2010:3424-3431
[16] Chua T S,Tang J,Hong R,et al.NUS-WIDE:a real-world web image database from National University of Singapore[C]∥Proceedings of the ACM International Conference on Image and Video Retrieval.2009:1-9
[17] Bingham E,Mannila H.Random projection in dimensionality reduction:applications to Image and text data[C]∥Proceedings of the International Conference on Knowledge Discovery and Data Mining.2001:245-250
[18] Bentley J L.Multidimensional binary search trees used for associative searching[J].Commun ACM,1975,18(9):509-517
[19] Liu W,Wang J,Ji R,et al.Supervised hashing with kernels[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2012:2074-2081
[20] LeCun Y,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324
[21] Krizhevsky A,Sutskever I,Hinton G E.Imagenet classification with deep convolutional neural networks[C]∥Proceedings of Advances in Neural Information Processing Systems.2012:1106-1114
[22] Hinton G E,Srivastava N,Krizhevsky A,et al.Improving neural networks by preventing co-adaptation of feature detectors[J].arXiv preprint,arXiv:1207.0580,2012
[23] Fukushima K,Miyake S.Neocognitron:A new algorithm forpattern recognition tolerant of deformations and shifts in position[J].Pattern recognition,1982,15(6):455-469
[24] LeCun B B,Denker J S,Henderson D,et al.Handwritten digit recognition with a back-propagation network[C]∥Proceedings of Advances in Neural Information Processing Systems.1990
[25] LeCun Y,Boser B,Denker J S,et al.Backpropagation applied to handwritten zip code recognition[J].Neural Computation,1989,1(4):541-551
[26] Hubel D H,Wiesel T N.Receptive fields,binocular interaction and functional architecture in the cat’s visual cortex[J].The Journal of physiology,1962,160(1):106-154
[27] Lin G,Shen C,Suter D,et al.A general two-step approach tolearning-based hashing[C]∥Proceedings of the IEEE Confe-rence on Computer Vision and Pattern Recognition.2013
[28] Kulis B,Darrell T.Learning to Hash with Binary Reconstructive Embeddings[C]∥Proceedings of the Advances in Neural Information Processing Systems.2009,22:1042-1050
[29] Liu W,Wang J,Kumar S,et al.Hashing with graphs[C]∥Proceedings of the International Conference on Machine Learning.2011:1-8
[30] Norouzi M,Blei D M.Minimal loss hashing for compact binary codes[C]∥Proceedings of the International Conference on Machine Learning.2011:353-360
[31] Yu C N J,Joachims T.Learning structural svms with latent va-riables[C]∥Proceedings of the International Conference on Machine Learning.2009:1169-1176
[32] Ciaccia P,Patella M,Zezula P.M-tree:An Efficient AccessMethod for Similarity Search in Metric Spaces[C]∥Proceedings of the International Conference on Very Large Data Bases.1997
[33] Beygelzimer A,Kakade S,Langford J.Cover trees for nearest neighbor[C]∥Proceedings of the International Conference on Machine Learning.2006:97-104
[34] Uhlmann J K.Satisfying general proximity/similarity querieswith metric trees[J].Information Processing Letters,1991,40(4):175-179
[35] Krizhevsky A,Hinton G.Learning multiple layers of featuresfrom tiny images[D].University of Toronto,2009
[36] Oliva A,Torralba A.Modeling the shape of the scene:A holistic representation of the spatial envelope[J].International Journal of Computer Vision,2001,42(3):145-175
[37] Wang J,Yang J,Yu K,et al.Locality-constrained linear coding for image classification[C]∥Proceedings of the IEEE Confe-rence on Computer Vision and Pattern Recognition.2010:3360-3367
[38] Lowe D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision,2004,60(2):91-110
[39] Csurka G,Dance C,Fan L,et al.Visual categorization with bags of keypoints[J].Proceedings of the European Conference on Computer Vision,2004,1(1-22):1-2
[40] Jia Y,Shelhamer E,Donahue J,et al.Caffe:Convolutional architecture for fast feature embedding.arXiv preprint,arXiv:1408.5093,2014
[41] Indyk P,Motwani R.Approximate nearest neighbors:towards removing the curse of dimensionality[C]∥Proceedings of the Symposium on Theory of Computing.1998:604-613
[42] Johnson W,Lindenstrauss J.Extensions of Lipschitz mappings into a Hilbert space[J].Contemporary Mathematics,1984,26:189-206

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 雷丽晖,王静. 可能性测度下的LTL模型检测并行化研究[J]. 计算机科学, 2018, 45(4): 71 -75, 88 .
[2] 夏庆勋,庄毅. 一种基于局部性原理的远程验证机制[J]. 计算机科学, 2018, 45(4): 148 -151, 162 .
[3] 厉柏伸,李领治,孙涌,朱艳琴. 基于伪梯度提升决策树的内网防御算法[J]. 计算机科学, 2018, 45(4): 157 -162 .
[4] 王欢,张云峰,张艳. 一种基于CFDs规则的修复序列快速判定方法[J]. 计算机科学, 2018, 45(3): 311 -316 .
[5] 孙启,金燕,何琨,徐凌轩. 用于求解混合车辆路径问题的混合进化算法[J]. 计算机科学, 2018, 45(4): 76 -82 .
[6] 张佳男,肖鸣宇. 带权混合支配问题的近似算法研究[J]. 计算机科学, 2018, 45(4): 83 -88 .
[7] 伍建辉,黄中祥,李武,吴健辉,彭鑫,张生. 城市道路建设时序决策的鲁棒优化[J]. 计算机科学, 2018, 45(4): 89 -93 .
[8] 刘琴. 计算机取证过程中基于约束的数据质量问题研究[J]. 计算机科学, 2018, 45(4): 169 -172 .
[9] 钟菲,杨斌. 基于主成分分析网络的车牌检测方法[J]. 计算机科学, 2018, 45(3): 268 -273 .
[10] 史雯隽,武继刚,罗裕春. 针对移动云计算任务迁移的快速高效调度算法[J]. 计算机科学, 2018, 45(4): 94 -99, 116 .