对抗性自动编码网络在高光谱异常检测中的应用

doi:10.11947/j.AGCS.2023.20220635

摘要/Abstract

摘要： 自动编码器（autoencoder，AE）是一种典型的生成模型。由于它具有简单的学习过程、良好的收敛能力和无监督的特性而得到了广泛的应用。AE的目标函数仅是输入输出之间的重构误差。为了提高其性能，提出了对抗性自动编码器（adversarial autoencoder，AAE），可以为原始的AE网络提供变分推理输出。本文回顾有关无监督和半监督的AAE模型在高光谱异常检测（hyperspectral anomaly detection，HAD）中的应用。除了在隐层空间中使用对抗性学习外，还可以通过在编码器的输入和解码器的输出之间添加对抗性学习来提高AAE的性能；通过这种方式，改进后的AAE网络可以更专注于学习数据分布而不仅是点对点的数值重建。试验结果表明，利用这些深度学习模型完成HAD任务的想法超越了传统HAD方法的概念，显著提高了检测性能。

关键词: 高光谱影像, 异常检测, 自动编码器, 对抗性自动编码器, 对抗学习

Abstract: Autoencoder (AE) is a typical generative model. It has been widely used due to its simple learning process, good ability for convergence, and unsupervised nature. To improve the performance of AE whose objective function is merely input-output reconstruction error, adversarial autoencoder (AAE) has been proposed, which can provide variational inference to the network output. This paper reviews the use of unsupervised and semisupervised AAE in hyperspectral anomaly detection (HAD). The performance of AAE can be improved by adding adversarial learning between the input of the encoder and the output of the decoder, in addition to the adversarial learning in the latent space in the original AAE. In this way, the network can focus more on learning data distribution rather than point-to-point data reconstruction. The idea of using these deep learning models is beyond the concept of traditional HAD methods, and can significantly improve the detection performance, as demonstrated by real data experiments.

Key words: hyperspectral Imagery, anomaly detection, autoencoder, adversarial autoencoder, adversarial learning

中图分类号:

P237

杜谦, 谢卫莹. 对抗性自动编码网络在高光谱异常检测中的应用[J]. 测绘学报, 2023, 52(7): 1105-1114.

DU Qian, XIE Weiying. Adversarial autoencoder for hyperspectral anomaly detection[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(7): 1105-1114.

参考文献

[1] 童庆禧, 张兵, 张立福. 中国高光谱遥感的前沿进展[J]. 遥感学报, 2016, 20(5):689-707. TONG Qingxi, ZHANG Bing, ZHANG Lifu. Current progress of hyperspectral remote sensing in China[J]. Journal of Remote Sensing, 2016, 20(5):689-707.
[2] WANG Qi, LIN Jianzhe, YUAN Yuan. Salient band selection for hyperspectral image classification via manifold ranking[J]. IEEE Transactions on Neural Networks and Learning Systems, 2016, 27(6):1279-1289.
[3] 贺霖, 潘泉, 邸韡, 等. 高光谱图像目标检测研究进展[J]. 电子学报, 2009, 37(9):2016-2024. HE Lin, PAN Quan,DI Wei, et al. Research advance on target detection for hyperspectral imagery[J]. Acta Electronica Sinica, 2009, 37(9):2016-2024.
[4] LI Lu, LI Wei, DU Qian, et al. Low-rank and sparse decomposition with mixture of Gaussian for hyperspectral anomaly detection[J]. IEEE Transactions on Cybernetics, 2021, 51(9):4363-4372.
[5] LI Shutao, ZHANG Kunzhong, DUAN Puhong, et al. Hyperspectral anomaly detection with kernel isolation forest[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(1):319-329.
[6] REED I S, YU X. Adaptive multiple-band CFAR detection of an optical pattern with unknown spectral distribution[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1990, 38(10):1760-1770.
[7] MOLERO J M, GARZÓN E M, GARCÍA I, et al. Analysis and optimizations of global and local versions of the RX algorithm for anomaly detection in hyperspectral data[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2013, 6(2):801-814.
[8] KWON H, NASRABADI N M. Kernel RX-algorithm:a nonlinear anomaly detector for hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(2):388-397.
[9] LI Wei, DU Qian. Collaborative representation for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(3):1463-1474.
[10] LI Jiayi, ZHANG Hongyan, ZHANG Liangpei, et al. Hyperspectral anomaly detection by the use of background joint sparse representation[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015, 8(6):2523-2533.
[11] WANG Rong, HU Haojie, HE Fang, et al. Self-weighted collaborative representation for hyperspectral anomaly detection[J]. Signal Processing, 2020, 177:107718.
[12] ZHANG Yuxiang, DU Bo, ZHANG Liangpei, et al. A low-rank and sparse matrix decomposition-based mahalanobis distance method for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(3):1376-1389.
[13] KANG Xudong, ZHANG Xiangping, LI Shutao, et al. Hyperspectral anomaly detection with attribute and edge-preserving filters[J]. IEEE Transactions on Geoscience and Remote Sensing, 2017, 55(10):5600-5611.
[14] XIE Weiying, JIANG Tao, LI Yunsong, et al. Structure tensor and guided filtering-based algorithm for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(7):4218-4230.
[15] TAO Ran, ZHAO Xudong, LI Wei, et al. Hyperspectral anomaly detection by fractional Fourier entropy[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019, 12(12):4920-4929.
[16] HUANG Zhihong, FANG Leyuan, LI Shutao. Subpixel-pixel-superpixel guided fusion for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(9):5998-6007.
[17] CHANG Shizhen, DU Bo, ZHANG Liangpei. A subspace selection-based discriminative forest method for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(6):4033-4046.
[18] SCHLEGL T, SEEBÖCK P, WALDSTEIN S M, et al. FAnoGAN:fast unsupervised anomaly detection with generative adversarial networks[J]. Medical Image Analysis, 2019, 54:30-44.
[19] CAO V L, NICOLAU M, MCDERMOTT J. Learning neural representations for network anomaly detection[J]. IEEE Transactions on Cybernetics, 2019, 49(8):3074-3087.
[20] 范潇杰,陈振安.基于LSTM和多种异常定义的卫星异常检测方法[J].测控技术,2021,40(11):78-87,95. FAN Xiaojie,CHEN Zhen'an. Satellite anomaly detection system based on LSTM and various anomaly definitions[J]. Measurement & Control Technology, 2021, 40(11)78-87, 95.
[21] KIRAN B, THOMAS D, PARAKKAL R. An overview of deep learning based methods for unsupervised and semi-supervised anomaly detection in videos[J]. Journal of Imaging, 2018, 4(2):36.
[22] ZARE A, JIAO C, GLENN T. Discriminative multiple instance hyperspectral target characterization[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(10):2342-2354.
[23] HINTON G E, ZEMEL R S. Autoencoders, minimum description length and Helmholtz free energy[J]. Advances in Neural Information Processing Systems, 1994, 6:3-10.
[24] HONG Y, HWANG U, YOO J, et al. How generative adversarial networks and their variants work[J]. ACM Computing Surveys, 2020, 52(1):1-43.
[25] XIE Weiying, LIU Baozhu, LI Yunsong, et al. Spectral adversarial feature learning for anomaly detection in hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(4):2352-2365.
[26] LU Xiaoqiang, ZHANG Wuxia, HUANG Ju. Exploiting embedding manifold of autoencoders for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(3):1527-1537.
[27] LEI Jie, FANG Shuo, XIE Weiying, et al. Discriminative reconstruction for hyperspectral anomaly detection with spectral learning[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(10):7406-7417.
[28] WANG Shaoyu, WANG Xinyu, ZHANG Liangpei, et al. Auto-AD:autonomous hyperspectral anomaly detection network based on fully convolutional autoencoder[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:1-14.
[29] XIANG P, ALI S, JUNG S K, et al. Hyperspectral anomaly detection with guided autoencoder[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:1-18.
[30] WEI Jie, ZHANG Jingfa, XU Yang, et al. Hyperspectral anomaly detection based on graph regularized variational autoencoder[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19:1-5.
[31] JIANG Tao, LI Yunsong, XIE Weiying, et al. Discriminative reconstruction constrained generative adversarial network for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(7):4666-4679.
[32] LI Wei, WU Guodong, DU Qian. Transferred deep learning for anomaly detection in hyperspectral imagery[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(5):597-601.
[33] SONG Shangzhen, ZHOU Huixin, YANG Yixin, et al. Hyperspectral anomaly detection via convolutional neural network and low rank with density-based clustering[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019, 12(9):3637-3649.
[34] YANG Xu, DENG Cheng, ZHENG Feng, et al. Deep spectral clustering using dual autoencoder network[C]//Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach:IEEE,2020:4061-4070.
[35] JIANG Kai, XIE Weiying, LI Yunsong, et al. Semisupervised spectral learning with generative adversarial network for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(7):5224-5236.
[36] XIE Weiying, LIU Baozhu, LI Yunsong, et al. Autoencoder and adversarial-learning-based semisupervised background estimation for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(8):5416-5427.
[37] BEGGEL L, PFEIFFER M, BISCHL B. Robust anomaly detection in images using adversarial autoencoders[EB/OL].[2019.06-03].https://arxiv.org/abs/1901.06355.
[38] Goodfellow I. et al.Generative adversarial nets[C]//Proceedings of 2014 Advances Inneural Information Processing Systems. Montreal, 2014:SPRINGER, 2014,2672-2680.
[39] MAKHZANI A, SHLENS J, JAITLY N, et al. Adversarial autoencoders[EB/OL].[2015-05-06].https://arxiv.org/abs/1511.05644.
[40] LEVEAU V, JOLY A. Adversarial autoencoders for novelty detection[C]//Proceedings of International Conference on Learning Representations.Toulon:ICLR, 2017.
[41] VU H S, UETA D, HASHIMOTO K, et al. Anomaly detection with adversarial dual autoencoders[EB/OL].[2019-06-23].https://arxiv.org/abs/1902.06924.
[42] ZHONG Jiaping, XIE Weiying, LI Yunsong, et al. Characterization of background-anomaly separability with generative adversarial network for hyperspectral anomaly detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(7):6017-6028.