基于松弛Hadamard矩阵的多模态融合哈希方法
Multimodal Fusion Hash Learning Method Based on Relaxed Hadamard Matrix
查看参考文献36篇
|
文摘
|
哈希作为一种有效的数据表征技术,已经在应对爆炸式增长的多媒体数据中扮演了重要的角色.它由于低存储和高效率的优势,在多媒体检索领域受到了越来越多的关注.目前多模态哈希学习方法在多媒体检索任务中得到了较好的研究和发展.然而,多数的方法通过编码特征的内积重构成对相似度来保持原始数据的结构信息,但是带来较复杂的优化问题.此外一些模型缺乏判别性使得检索性能的提升受到限制.为了克服上述问题,本文提出一种新型的多模态融合哈希方法,在类别信息的监督下利用Hadamard矩阵为数据生成目标编码,通过松弛严格的二值约束增大类间的间隔,同时采用图嵌入的方式促进类内的紧凑性.本文提出的方法既保证了模型具有很好的判别能力也简化了优化过程.在3个公开数据集上的实验结果表明,本文提出的方法在多媒体数据检索中是非常有效的,平均性能上相比最优的对比方法提高了8.47%. |
|
其他语种文摘
|
Hashing, as an effective data representation technology, has played an important role in dealing with the explosive growth of multimedia data. Due to the advantages of its low storage and high efficiency, it has received more and more attention in the field of multimedia retrieval. At present, multi-modal hashing methods have been well researched and developed in multimedia retrieval tasks. However, most of these methods usually use the inner product of hashing features to reconstruct larger pairwise similarity, aiming to preserve the structural information of the original data, which will bring more complex optimization problems. Besides, some models lack discriminant ability, which leads to limitations in the improvement of retrieval performance. In order to overcome the above-mentioned problems, this paper proposes a new multimodal fusion hashing method. Under the supervision of category information, Hadamard matrix is used to generate target codes for data, and the margin between categories is increased by relaxing strict binary constraints. At the same time, the graph embedding approach is used to promote compactness within the class. The proposed method in this paper not only ensures the strong discriminative ability of the model, but also simplifies the optimization process. The experimental results on three public datasets show that the method proposed in this paper is very effective in multimedia data retrieval, and the average performance is 8.47% higher than that of the optimal comparison method. |
|
来源
|
电子学报
,2022,50(4):909-920 【核心库】
|
|
DOI
|
10.12263/DZXB.20210760
|
|
关键词
|
哈希学习
;
多模态融合
;
Hadamard矩阵
;
多媒体检索
;
哈希中心
|
|
地址
|
1.
郑州轻工业大学计算机与通信工程学院, 河南, 郑州, 450000
2.
江南大学计算机与人工智能学院, 江苏, 无锡, 214000
|
|
语种
|
中文 |
|
文献类型
|
研究性论文 |
|
ISSN
|
0372-2112 |
|
学科
|
自动化技术、计算机技术 |
|
基金
|
河南省科技攻关项目
;
郑州轻工业大学博士科研启动基金
;
国家自然科学基金
|
|
文献收藏号
|
CSCD:7190617
|
参考文献 共
36
共2页
|
|
1.
Wang J.
Hashing for similarity search: A survey,2014
|
CSCD被引
1
次
|
|
|
|
|
2.
李志欣. 融合两级相似度的跨媒体图像文本检索.
电子学报,2021,49(2):268-274
|
CSCD被引
10
次
|
|
|
|
|
3.
李武军. 大数据哈希学习:现状与趋势.
科学通报,2015,60(5):485-490
|
CSCD被引
31
次
|
|
|
|
|
4.
高文. "存得下,查得快"拥抱多媒体大数据时代.
创新科技,2013,25(6):7
|
CSCD被引
1
次
|
|
|
|
|
5.
刘昊淼. 基于离散优化的哈希编码学习方法.
计算机学报,2019,42(5):1149-1160
|
CSCD被引
4
次
|
|
|
|
|
6.
Gong Y. Iterative quantization: A procrustean approach to learning binary codes for large-scale image retrieval.
IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,35(12):2916-2929
|
CSCD被引
113
次
|
|
|
|
|
7.
Kulis B. Kernelized locality-sensitive hashing for scalable image search.
Proceedings of 12th International Conference on Computer Vision,2009:2130-2137
|
CSCD被引
1
次
|
|
|
|
|
8.
Ji R. Toward optimal manifold hashing via discrete locally linear embedding.
IEEE Transactions on Image Processing,2017,26(11):5411-5420
|
CSCD被引
4
次
|
|
|
|
|
9.
Koutaki G. Hadamard coding for supervised discrete hashing.
IEEE Transactions on Image Processing,2018,27(11):5378-5392
|
CSCD被引
2
次
|
|
|
|
|
10.
Jin L. Weakly-supervised image hashing through masked visual-semantic graph-based reasoning.
Proceedings of the 28th ACM International Conference on Multimedia,2020:916-924
|
CSCD被引
2
次
|
|
|
|
|
11.
Li Z. Weakly-supervised semantic guided hashing for social image retrieval.
International Journal of Computer Vision,2020,128:2265-2278
|
CSCD被引
9
次
|
|
|
|
|
12.
姚涛. 基于映射字典学习的跨模态哈希检索.
自动化学报,2018,44(8):1475-1485
|
CSCD被引
4
次
|
|
|
|
|
13.
Wang D. Label consistent matrix factorization hashing for large-scale cross-modal similarity search.
IEEE Transactions on Pattern Analysis and Machine Intelligence,2018,41(10):2466-2479
|
CSCD被引
6
次
|
|
|
|
|
14.
Wang D. Robust and flexible discrete hashing for cross-modal similarity search.
IEEE Transactions on Circuits and Systems for Video Technology,2017,28(10):2703-2715
|
CSCD被引
5
次
|
|
|
|
|
15.
刘昊鑫. 联合哈希特征和分类器学习的跨模态检索算法.
模式识别与人工智能,2020,33(2):160-165
|
CSCD被引
1
次
|
|
|
|
|
16.
Liu H. Supervised matrix factorization for cross-modality hashing.
Proceedings of International Joint Conference on Artificial Intelligence,2016:1767-1773
|
CSCD被引
1
次
|
|
|
|
|
17.
Lin Z. Semantics-preserving hashing for cross-view retrieval.
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2015:3864-3872
|
CSCD被引
3
次
|
|
|
|
|
18.
王锦荟.
基于知识蒸馏的跨模态哈希,2021
|
CSCD被引
1
次
|
|
|
|
|
19.
Li Z. Deep collaborative embedding for social image understanding.
IEEE Transactions on Pattern Analysis and Machine Intelligence,2018,41(9):2070-2083
|
CSCD被引
21
次
|
|
|
|
|
20.
Song J. Effective multiple feature hashing for large scale near duplicate video retrieval.
IEEE Transaction on Multimedia,2013,15(8):1997-2008
|
CSCD被引
15
次
|
|
|
|
|
了解气象卫星更多信息，请访问