[1]王德文,魏波涛.基于孪生变分自编码器的小样本图像分类方法[J].智能系统学报,2021,16(2):254-262.[doi:10.11992/tis.201906022]
WANG Dewen,WEI Botao.A small-sample image classification method based on a Siamese variational auto-encoder[J].CAAI Transactions on Intelligent Systems,2021,16(2):254-262.[doi:10.11992/tis.201906022]
点击复制
WANG Dewen,WEI Botao.A small-sample image classification method based on a Siamese variational auto-encoder[J].CAAI Transactions on Intelligent Systems,2021,16(2):254-262.[doi:10.11992/tis.201906022]
基于孪生变分自编码器的小样本图像分类方法
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
16
期数:
2021年第2期
页码:
254-262
栏目:
学术论文—机器感知与模式识别
出版日期:
2021-03-05
- Title:
- A small-sample image classification method based on a Siamese variational auto-encoder
- Keywords:
- small sample; variational auto-encoder (VAE); siamese network; image recognition; over-fitting; eigenvector; deep learning; data augmentation
- 分类号:
- TP183
- 摘要:
- 当前深度学习大都基于大量数据通过构建深层次的网络实现自动识别，但在很多场景中难以获得大量的样本数据。针对这一问题，提出一种基于孪生变分自编码器(siamese variational auto-encoder，S-VAE)的小样本图像分类方法。通过变分自编码器提取原始训练数据的高层语义特征，然后由两个训练好的变分自编码器的编码器部分组建孪生网络的输入结构，最后通过分类器对样本进行识别。变分自编码器可以解决样本数据量少带来的过拟合问题，孪生网络的结构增加了样本数量较少的情况下的训练次数。在Omniglot数据集上进行的实验结果表明：本方法与原始孪生神经网络相比正确率平均提高了3.1%，模型收敛速度更快，证明了孪生变分自编码器能够较好地完成小样本数据分类任务。
- Abstract:
- Currently, most deep learning is based on the use of large amounts of data and the construction of a deep network to achieve automatic recognition, but it is difficult to obtain a large amount of sample data in many scenarios. To solve this problem, we propose the use of a small-sample image classification method based on a Siamese variational auto-encoder (S-VAE). First, the high-level semantic features of the original training data are extracted by the variational auto-encoder. Then, the input structure of the Siamese network is constructed by the encoders of two trained variational auto-encoders. Lastly, the samples are identified by the classifier. The variational auto-encoder can solve the over-fitting problem caused by small amounts of sample data, and the structure of the Siamese network increases the training number when the sample size is small. The experimental results on the Omniglot dataset show that this method has a 3.1% average improvement in accuracy compared with the original Siamese network. The model convergence speed is also faster, which proves that the S-VAE can better complete the classification task when using small-sample data.
- 参考文献/References:
-
[1] 胡越, 罗东阳, 花奎, 等. 关于深度学习的综述与讨论[J]. 智能系统学报, 2019, 14(1):1-19
HU Yue, LUO Dongyang, HUA Kui, et al. Overview on deep learning[J]. CAAI transactions on intelligent systems, 2019, 14(1):1-19
[2] 马世龙, 乌尼日其其格, 李小平. 大数据与深度学习综述[J]. 智能系统学报, 2016, 11(6):728-742
MA Shilong, WUNIRI Qiqige, LI Xiaoping. Deep learning with big data:state of the art and development[J]. CAAI transactions on intelligent systems, 2016, 11(6):728-742
[3] 王昊, 刘高军, 段建勇, 等. 基于特征自学习的交通模式识别研究[J]. 哈尔滨工程大学学报, 2019, 40(2):354-358
WANG Hao, LIU GaoJun, DUAN Jianyong, et al. Transportation mode detection based on self-learning of features[J]. Journal of Harbin Engineering University, 2019, 40(2):354-358
[4] 张程熠, 唐雅洁, 李永杰, 等. 适用于小样本的神经网络光伏预测方法[J]. 电力自动化设备, 2017, 37(1):101-106, 111
ZHANG Chengyi, TANG Yajie, LI Yongjie, et al. Photovoltaic power forecast based on neural network with a small number of samples[J]. Electric power automation Equipment, 2017, 37(1):101-106, 111
[5] 洪雁飞, 魏本征, 刘川, 等. 基于深度学习的椎间孔狭窄自动多分级研究[J]. 智能系统学报, 2019, 14(4):1-9
HONG Feiyan, WEI Benzheng, LIU Chuan, et al. Deep learning based automatic multi-classification algorithm for intervertebral foraminal stenosis[J]. CAAI transactions on intelligent systems, 2019, 14(4):1-9
[6] 王翔, 胡学钢. 高维小样本分类问题中特征选择研究综述[J]. 计算机应用, 2017, 37(9):2433-2438, 2448
WANG Xiang, HU Xuegang. Overview on feature selection in high-dimensional and small-sample-size classification[J]. Journal of computer applications, 2017, 37(9):2433-2438, 2448
[7] LI Feifei, FERGUS R, PERONA P. One-shot learning of object categories[J]. IEEE trans pattern anal mach intell, 2006, 28(4):594-611.
[8] LAKE B M, SALAKHUTDINOV R, TENENBAUM J B. Human-level concept learning through probabilistic program induction[J]. Science, 2015, 350(6266):1332-1338.
[9] 宋丽丽. 迁移度量学习行人再识别算法[J]. 计算机工程与应用, 2019, 55(20):170-176, 201
SONG Lili. Transfer metric learning for person re-identification[J]. Computer engineering and applications, 2019, 55(20):170-176, 201
[10] 任俊, 胡晓峰, 朱丰. 基于深度学习特征迁移的装备体系效能预测[J]. 系统工程与电子技术, 2017, 39(12):2745-2749
REN Jun, HU Xiaofeng, ZHU Feng. Effectiveness prediction of weapon equipment system-of-systems based on deep learning feature transf[J]. Systems engineering and electronics, 2017, 39(12):2745-2749
[11] 谭本东, 杨军, 赖秋频, 等. 基于改进CGAN的电力系统暂态稳定评估样本增强方法[J]. 电力系统自动化, 2019, 43(1):149-160
TAN Bendong, YANG Jun, LAI QiuPin, et al. Data augment method for power system transient stability assessment based on improved conditional generative adversarial network[J]. Automation of electric power systems, 2019, 43(1):149-160
[12] 王建敏, 吴云洁. 基于聚类云模型的小样本数据可信度评估[J]. 系统仿真学报, 2019, 31(7):1263-1271
WANG Jianmin, WU Yunjie. Credibility evaluation method of small sample data based on cluster cloud model[J]. Journal of system simulation, 2019, 31(7):1263-1271
[13] 杨懿男, 齐林海, 王红, 等. 基于生成对抗网络的小样本数据生成技术研究[J]. 电力建设, 2019, 40(5):71-77
YANG Yinan, QI Linhai, WANG Hong, et al. Research on generation technology of small sample data based on generative adversarial network[J]. Electric power construction, 2019, 40(5):71-77
[14] 韩冬, 马进, 贺仁睦. 基于Bootstrap的实测负荷模型参数优选[J]. 电工技术学报, 2012, 27(8):141-146
HAN Dong, MA Jin, HE Renmu. Parameter optimization of measurement-based load model based on bootstrap[J]. Transactions of China electrotechnical society, 2012, 27(8):141-146
[15] 马晓, 张番栋, 封举富. 基于深度学习特征的稀疏表示的人脸识别方法[J]. 智能系统学报, 2016, 11(3):279-286
MA Xiao, ZHANG Fandong, FENG Jufu. Sparse representation via deep learning features based face recognition method[J]. Transactions of China electrotechnical society, 2016, 11(3):279-286
[16] 马忠丽, 刘权勇, 武凌羽, 等. 一种基于联合表示的图像分类方法[J]. 智能系统学报, 2018, 13(2):220-226
MA Zhongli, LIU Quanyong, WU Lingyu, et al. Syncretic representation method for image classification[J]. CAAI transactions on intelligent systems, 2018, 13(2):220-226
[17] 赵春晖, 齐滨, Eunseog Youn. 基于蒙特卡罗特征降维算法的小样本高光谱图像分类[J]. 红外与毫米波学报, 2013, 32(1):62-67
ZHAO Chun Hui, QI Bin, EUNSEOG Youn. Hyperspectral image classification based on Monte Carlo feature reduction method[J]. Journal of infrared and millimeter waves, 2013, 32(1):62-67
[18] CHOPRA S, HADSELL R, LECUN Y. Learning a similarity metric discriminatively, with application to face verification[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), San Diego, USA, 2005:539-546.
[19] KOCH G, ZEMEL R, SALAKHUTDINOV R. Siamese neural networks for one-shot image recognition[C]//Proc of the ICML Deep Learning Workshop, Lille, France 2015.
[20] KINGMA D P, WELLING M. Auto-encoding variational bayes[C]//International Conference on Learning Representations,[S.l.], 2014.
[21] LIU Guojun, LIUYang, GUO Maozu, et al. Variational inference with Gaussian mixture model and householder flow[J]. Neural networks, 2019, 109:43-55.
[22] 宋辉, 代杰杰, 张卫东, 等. 基于变分贝叶斯自编码器的局部放电数据匹配方法[J]. 中国电机工程学报, 2018, 38(19):5869-5877, 5945
SONG Hui, DAI Jiejie, ZHANG Weidong, et al. A data matching method of partial discharge data based on auto-encoding briational Bayes[J]. Proceedings of the CSEE, 2018, 38(19):5869-5877, 5945
[23] Lu Guangquan, Zhao Xishun, Yin Jian, et al. Multi-task learning using variational auto-Encoder for sentiment classification[J]. Pattern recognition letters, 2018.
[24] ZAKHAROV N, SU H, ZHU J, et al. Towards controllable image descriptions with semi-supervised VAE[J]. Journal of visual communication and image representation, 2019, 63:102574.
[25] VINYALS O, BLUNDELL C, LILLICRAP T, et al. Matching networks for one shot learning[C]//Proc of the Advances in Neural Information Processing Systems, 2016:3630-3638
[26] LAKE B M, SALAKHUTDINOV R, TENENBAUM J B. One-shot learning by inverting a compositional causal process[C]//International Conference on Neural Information Processing Systems. Curran Associates Inc 2013.
[27] ZHANG Ling, lIU Jun, LUO Minnan, et al. Hauptmann, Scheduled sampling for one-shot learning via matching network[J]. Pattern Recongnition, 2019, 96:106962.
- 相似文献/References:
-
[1]马晓,张番栋,封举富.基于深度学习特征的稀疏表示的人脸识别方法[J].智能系统学报,2016,11(3):279.[doi:10.11992/tis.201603026]
MA Xiao,ZHANG Fandong,FENG Jufu.Sparse representation via deep learning features based face recognition method[J].CAAI Transactions on Intelligent Systems,2016,11():279.[doi:10.11992/tis.201603026]
[2]马忠丽,刘权勇,武凌羽,等.一种基于联合表示的图像分类方法[J].智能系统学报,2018,13(2):220.[doi:10.11992/tis.201611036]
MA Zhongli,LIU Quanyong,WU Lingyu,et al.Syncretic representation method for image classification[J].CAAI Transactions on Intelligent Systems,2018,13():220.[doi:10.11992/tis.201611036]
[3]狄岚,矫慧文,梁久祯.稀疏综合字典学习的小样本人脸识别[J].智能系统学报,2021,16(2):218.[doi:10.11992/tis.201910028]
DI Lan,JIAO Huiwen,LIANG Jiuzhen.Sparse comprehensive dictionary learning for small-sample face recognition[J].CAAI Transactions on Intelligent Systems,2021,16():218.[doi:10.11992/tis.201910028]
[4]毛莺池,唐江红,王静,等.基于Faster R-CNN的多任务增强裂缝图像检测方法[J].智能系统学报,2021,16(2):286.[doi:10.11992/tis.201910004]
MAO Yingchi,TANG Jianghong,WANG Jing,et al.Multi-task enhanced dam crack image detection based on Faster R-CNN[J].CAAI Transactions on Intelligent Systems,2021,16():286.[doi:10.11992/tis.201910004]
[5]张冀,曹艺,王亚茹,等.融合VAE和StackGAN的零样本图像分类方法[J].智能系统学报,2022,17(3):593.[doi:10.11992/tis.202107012]
ZHANG Ji,CAO Yi,WANG Yaru,et al.Zero-shot image classification method combining VAE and StackGAN[J].CAAI Transactions on Intelligent Systems,2022,17():593.[doi:10.11992/tis.202107012]
[6]邢天祎,郭茂祖,陈加栋,等.基于空时对抗变分自编码器的人群异常行为检测[J].智能系统学报,2023,18(5):994.[doi:10.11992/tis.202303002]
XING Tianyi,GUO Maozu,CHEN Jiadong,et al.Detection of abnormal crowd behavior based on spatial-temporal adversarial variational autoencoder[J].CAAI Transactions on Intelligent Systems,2023,18():994.[doi:10.11992/tis.202303002]
[7]毕晓君,毛亚菲.基于监督对比学习的小样本甲骨文字识别[J].智能系统学报,2024,19(1):106.[doi:10.11992/tis.202309008]
BI Xiaojun,MAO Yafei.Few-shot oracle bone character recognition based on supervised contrastive learning[J].CAAI Transactions on Intelligent Systems,2024,19():106.[doi:10.11992/tis.202309008]
备注/Memo
收稿日期:2019-06-12。
基金项目:国家自然科学基金项目（51677072）
作者简介:王德文，副教授，博士，主要研究方向为人工智能与大数据。发表学术论文60余篇;魏波涛，硕士研究生，主要研究方向为人工智能与图像处理
通讯作者:魏波涛.E-mail:764387445@qq.com
更新日期/Last Update:
2021-04-25