基于显著图加权视觉语言模型的图像分类方法

doi:10.3969/j.issn.1000-3428.2015.03.039

摘要/Abstract

摘要： 传统基于视觉语言模型(VLM)的图像分类方法在参数估计阶段,通常采用最大似然估计的方式统计视觉单词的分布,忽略了图像背景噪声对该模型参数估计的影响。为此,提出一种新的图像分类方法。利用基于视觉注意的显著性检测算法提取图像中的显著区域和背景区域,构建的图像带有显著图标识的视觉文档,训练视觉单词的显著度权重和条件概率,并使用显著图加权视觉语言模型进行图像分类。实验结果表明,与传统VLM 等方法相比,该方法能有效克服图像背景噪声的影响,增强视觉单词的区分性,提高分类准确率。

关键词: 图像信息, 视觉语言模型, 图像分类, 背景区域, 显著图

Abstract: At the parameter estimation stage of the image classification method based on the traditional Visual Language Model(VLM),the distribution of visual words is usually analysed via maximum likelihood estimation,which ignores the adverse effect of image background noise. In view of the problem,an image classification method of weighted VLM for saliency map is put forward. The salient regions and background regions are extracted via saliency detection algorithm based on visual attention,the visual documents of images with salient labels are constructed,and the salient weights and conditional probability are estimated in the training phase. The images are classified with weighted VLM for saliency map. Experimental results show that,this method can effectively reduce the influence of image background noise,and enhances the discrimination performance of visual words,so as to improve the performance of image classification based on VLM.

Key words: image information, Visual Language Model(VLM), image classification, background region, saliency map

中图分类号:

TP391

王挺进,赵永威,李弼程. 基于显著图加权视觉语言模型的图像分类方法[J]. 计算机工程, doi: 10.3969/j.issn.1000-3428.2015.03.039.

WANG Tingjin,ZHAO Yongwei,LI Bicheng. Image Classification Method Based on Weighted Visual Language Model for Saliency Map[J]. Computer Engineering, doi: 10.3969/j.issn.1000-3428.2015.03.039.

http://www.ecice06.com/CN/Y2015/V41/I3/204

参考文献

参考文献 [ 1 ]　Gao Haolin,Dou Linhui,Chen Weijin. Image Classification with Bag-of-Words Model Based on Improved SIFT Algorithm [ C ]/ / Proceedings of the 9th Asian Control Conference. [S. l. ]:IEEE Press,2013:1-6. [ 2 ]　刘硕研,须　德. 一种基于上下文语义信息的图像块视觉单词生成算法[J]. 电子学报,2010,38(5):1156-1161. [ 3 ]　赵永威,郭志刚,李弼程,等. 基于随机化视觉词典组和上下文语义信息的目标检索方法[J]. 电子学报, 2012,40(12):2472-2480. [ 4 ]　胡正平,涂潇蕾. 多方向上下文特征结合空间金字塔模型的场景分类[J]. 信号处理,2011,27 (10): 1536-1542. [ 5 ]　张琳波,王春恒,肖柏华,等. 基于Bag-of-Phrases 的图像表示方法[J]. 自动化学报,2012,38(1):46-54. [ 6 ]　Jiang Yang, Meng Ji, Yuan Jin. Randomized Visual Phrases for Object Search [ C] / / Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. New York,USA:IEEE Press,2012:3100-3107. [ 7 ]　Wu Lei,Li Meng,Li Zi,et al. Visual Language Modeling for Image Classification[C]/ / Proceedings of International Workshop on Workshop on Multimedia Information Retrieval. Pairs,France:IEEE Press,2007:115-124. [ 8 ]　Wu Lei,Hu Yong,Li Meng,et al. Scale-invariant Visual Language Modeling for Object Categorization[J]. IEEE Transactions on Multimedia,2009,11(2):286-294. [ 9 ]　Pham T T, Maisonnasse L, Mulhem P, et al. Visual Language Model for Scene Recognition[C]/ / Proceedings of Singaporean-French Ipal Symposium. New York,USA: ACM Press,2009:76-85. [10]　Narayanaswamy S, Barbu A, Siskind J M. A Visual Language Model for Estimating Object Pose and Structure in a Generative Visual Domain[C]/ / Proceedings of IEEE International Conference on Robotics and Automation. Landon,UK:IEEE Press,2011:4854-4860. [11]　Li Minjin, Ma Wuyang. Visual Language Modeling for Image Classification: USA, US008126274B2 [ P ]. 2012-02-28. [12]　金　聪,刘金安,金枢炜. 基于视觉语言模型的图像语义挖掘研究[ J ]. 图书情报工作,2013,57 (5): 120-123. [13]　Liu Shi, Bai Xia. Discriminative Features for Image Classification and Retrieval [ J ]. Pattern Recognition Letters,2012,33(6):744-751. [14]　Rubner Y,Tomasi C, Guibas L J. The Earth Mover ’ s Distance as a Metric for Image Retrieval[J]. International Journal of Computer Vision,2000,40(2):99-121. [15]　Chen Tao,Kang Ya,Zhang Deng. Discriminative Soft Bagof- Visual Phrase for Mobile Landmark Recognition [J]. IEEE Transactions on Multimedia,2013,16(3):612-622. [16]　Yan Yupeng,Tian Xinmei,Yang Linjun,et al. Semanticspatial Matching for Image Classification[C]/ / Proceedings of IEEE International Conference on Multimedia and Expo. New York,USA:IEEE Press,2013:1-6. [17]　Kostavelis I,Nalpantidis L,Gasteratos A. Object Recognition Using Saliency Maps and HTM Learning[C]/ / Proceedings of IEEE International Conference on Imaging Systems and Techniques. Paris,France:IEEE Press,2012:528-532. [18]　Li Lin,Jiang Shan,Zha Zheng,et al. Partial-duplicate Image Retrieval via Saliency-guided Visually Matching[J]. IEEE Transactions on Multimedia,2013,20(3):13-23. [19]　Itti L,Koch C, Niebur E. A Model of Saliency-based Visual Attention for Rapid Scene Analysis [ J ]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(11):1254-1259. [20]　Harel J, Koch C, Perona P. Graph-based Visual Saliency [ C ] / / Proceedings of Advances in Neural Information Processing Systems. Washington D. C.,USA: IEEE Press,2006:545-552. [21]　马儒宁,涂小坡,丁军娣,等. 视觉显著性凸显目标的评价[J]. 自动化学报,2012,38(5):870-876. [22]　Li Feifei,Fergus R,Perona P. Learning Generative Visual Models from Few Training Examples: An Incremental Bayesian Approach Tested on 101 Object Categories[J]. Computer Vision and Image Understanding,2007,106(1): 59-70. 编辑　刘　冰

[1]	曹坪, 杨怀志, 薄一军, 尤嘉, 张淳杰, 李丹勇. 面向低质量裂缝图像的多知识蒸馏分类[J]. 计算机工程, 2023, 49(7): 204-213.
[2]	石进, 徐杨, 曹斌. 基于自适应三线性池化网络的细粒度图像分类[J]. 计算机工程, 2023, 49(5): 239-246,254.
[3]	韩彦岭, 沈思扬, 徐利军, 王静, 张云, 周汝雁. 面向深度学习图像分类的GPU并行方法研究[J]. 计算机工程, 2023, 49(1): 191-200.
[4]	陈珊, 孙仁诚, 邵峰晶, 隋毅. 基于胶囊网络的动态路由研究与改进[J]. 计算机工程, 2022, 48(5): 208-214.
[5]	袁单飞, 陈慈发, 董方敏. 基于多尺度分割的图像识别残差网络研究[J]. 计算机工程, 2022, 48(5): 258-262,271.
[6]	谭润, 叶武剑, 刘怡俊. 结合双语义数据增强与目标定位的细粒度图像分类[J]. 计算机工程, 2022, 48(2): 237-242,249.
[7]	刘鹏飞, 朱健晨, 万良易, 江波. 低功耗异构计算架构的高光谱遥感图像分类研究[J]. 计算机工程, 2022, 48(12): 9-15,23.
[8]	李伟, 黄鹤鸣, 武风英, 张会云. 基于深度多特征融合的自适应CNN图像分类算法[J]. 计算机工程, 2021, 47(9): 235-239,251.
[9]	陶洋, 鲍灵浪, 胡昊. 结合表示学习与嵌入子空间学习的降维方法[J]. 计算机工程, 2021, 47(6): 83-87,97.
[10]	赖妍菱, 石峻峰, 陈继鑫, 白汉利, 唐晓澜, 邓碧颖, 郑德生. 基于U-Net的对抗样本防御模型[J]. 计算机工程, 2021, 47(12): 163-170.
[11]	陈鑫华, 钱雪忠, 宋威. 基于轻量级特征融合卷积网络的图像分类算法[J]. 计算机工程, 2021, 47(11): 268-275.
[12]	肖经纬, 田军委, 王沁, 程希希, 王佳. 基于改进残差网络的果实病害分类方法[J]. 计算机工程, 2020, 46(9): 221-225.
[13]	李培媛, 黄迟. 基于CSPPNet与集成学习的人类蛋白质图像分类方法[J]. 计算机工程, 2020, 46(8): 235-242.
[14]	张驰名, 王庆凤, 刘志勤, 黄俊, 陈波, 付婕, 周莹. 基于深度学习的胸部常见病变诊断方法[J]. 计算机工程, 2020, 46(7): 306-311,320.
[15]	蔡凯, 李新福, 田学东. 基于视觉注意力分析的3D内容生成方法[J]. 计算机工程, 2020, 46(4): 266-272.

选择文件类型/文献管理软件名称

选择包含的内容