Abstract
Landslide recognition is widely used in natural disaster risk management. Traditional landslide recognition is mainly conducted by geologists, which is accurate but inefficient. This article introduces multiple instance learning (MIL) to perform automatic landslide recognition. An end-to-end deep convolutional neural network is proposed, referred to as Multiple Instance Learning–based Landslide classification (MILL). First, MILL uses a large-scale remote sensing image classification dataset to build pre-train networks for landslide feature extraction. Second, MILL extracts instances and assign instance labels without pixel-level annotations. Third, MILL uses a new channel attention–based MIL pooling function to map instance-level labels to bag-level label. We apply MIL to detect landslides in a loess area. Experimental results demonstrate that MILL is effective in identifying landslides in remote sensing images.
- Beliz Aksoy and Murat Ercanoglu. 2012. Landslide identification and classification by object-based image analysis and fuzzy logic: An example from the Azdavay region (Kastamonu, Turkey). Comput. Geosci. 38, 1 (2012), 87–98. Google ScholarDigital Library
- Dieu Tien Bui, Tran Anh Tuan, Harald Klempe, Biswajeet Pradhan, and Inge Revhaug. 2016. Spatial prediction models for shallow landslide hazards: A comparative assessment of the efficacy of support vector machines, artificial neural networks, kernel logistic regression, and logistic model tree. Landslides 13, 2 (2016), 361–378.Google ScholarCross Ref
- Alberto Cano. 2017. An ensemble approach to multi-view multi-instance learning. Knowl.-based Syst. 136 (2017), 46–57.Google Scholar
- Marc-André Carbonneau, Veronika Cheplygina, Eric Granger, and Ghyslain Gagnon. 2018. Multiple instance learning: A survey of problem characteristics and applications. Pattern Recog. 77 (2018), 329–353. Google ScholarDigital Library
- K. T. Chau, Y. L. Sze, M. K. Fung, W. Y. Wong, E. L. Fong, and L. C. P. Chan. 2004. Landslide hazard analysis for Hong Kong using landslide inventory and GIS. Comput. Geosci. 30, 4 (2004), 429–443.Google ScholarCross Ref
- Keren Dai, Zhenhong Li, Qiang Xu, Roland Burgmann, David G. Milledge, Roberto Tomas, Xuanmei Fan, Chaoying Zhao, Xiaojie Liu, Jianbing Peng et al. 2020. Entering the era of earth observation-based landslide warning systems: A novel and exciting framework. IEEE Geosci. Remote Sens. Mag. 8, 1 (2020), 136–153.Google Scholar
- Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. ImageNet: A large-scale hierarchical image database. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 248–255.Google ScholarCross Ref
- Mirco Galli, Francesca Ardizzone, Mauro Cardinali, Fausto Guzzetti, and Paola Reichenbach. 2008. Comparing landslide inventory maps. Geomorphology 94, 3–4 (2008), 268–289.Google ScholarCross Ref
- Stefano Luigi Gariano and Fausto Guzzetti. 2016. Landslides in a changing climate. Earth-sci. Rev. 162 (2016), 227–252.Google Scholar
- Omid Ghorbanzadeh, Thomas Blaschke, Khalil Gholamnia, Sansar Raj Meena, Dirk Tiede, and Jagannath Aryal. 2019. Evaluation of different machine learning methods and deep-learning convolutional neural networks for landslide detection. Remote Sens. 11, 2 (2019), 196.Google ScholarCross Ref
- Noel Gorelick, Matt Hancher, Mike Dixon, Simon Ilyushchenko, David Thau, and Rebecca Moore. 2017. Google earth engine: Planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 202 (2017), 18–27.Google ScholarCross Ref
- Fausto Guzzetti, Alberto Carrara, Mauro Cardinali, and Paola Reichenbach. 1999. Landslide hazard evaluation: A review of current techniques and their application in a multi-scale study, Central Italy. Geomorphology 31, 1–4 (1999), 181–216.Google ScholarCross Ref
- Fausto Guzzetti, Paola Reichenbach, Francesca Ardizzone, Mauro Cardinali, and Mirco Galli. 2006. Estimating the quality of landslide susceptibility models. Geomorphology 81, 1–2 (2006), 166–184.Google ScholarCross Ref
- Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 770–778.Google ScholarCross Ref
- Andrew Howard, Mark Sandler, Grace Chu, Liang-Chieh Chen, Bo Chen, Mingxing Tan, Weijun Wang, Yukun Zhu, Ruoming Pang, Vijay Vasudevan, Quoc V. Le, and Hartwig Adam. 2019. Searching for MobileNetV3. In Proceedings of the IEEE International Conference on Computer Vision. 1314–1324.Google ScholarCross Ref
- Jie Hu, Li Shen, and Gang Sun. 2018. Squeeze-and-excitation networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 7132–7141.Google ScholarCross Ref
- Maximilian Ilse, Jakub Tomczak, and Max Welling. 2018. Attention-based deep multiple instance learning. In Proceedings of the International Conference on Machine Learning. 2127–2136.Google Scholar
- Michel Jaboyedoff, Thierry Oppikofer, Antonio Abellán, Marc-Henri Derron, Alex Loye, Richard Metzger, and Andrea Pedrazzini. 2012. Use of LIDAR in landslide investigations: A review. Nat. Haz. 61, 1 (2012), 5–28.Google ScholarCross Ref
- Asako Kanezaki. 2018. Unsupervised image segmentation by backpropagation. In Proceedings of the IEEE International Conference On Acoustics, Speech And Signal Processing. 1543--1547.Google ScholarCross Ref
- Diederik P. Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. In Proceedings of the International Conference on Learning Representations.Google Scholar
- Zhongbin Li, Wenzhong Shi, Ping Lu, Lin Yan, Qunming Wang, and Zelang Miao. 2016. Landslide mapping from aerial photographs using change detection-based Markov random field. Remote Sens. Environ. 187 (2016), 76–90.Google ScholarCross Ref
- Zhili Li, Kai Xu, Jiafen Xie, Qi Bi, and Kun Qin. 2020. Deep multiple instance convolutional neural networks for learning robust scene representations. IEEE Trans. Geosci. Remote Sens. 58, 5 (2020), 3685–3702.Google ScholarCross Ref
- Mingzhe Liu, Xin Jiang, Yanhua Liu, Feixiang Zhao, and Helen Zhou. 2020. A semi-supervised convolutional transfer neural network for 3D pulmonary nodules detection. Neurocomputing 391 (2020), 199–209.Google ScholarCross Ref
- Mingxia Liu, Jun Zhang, Ehsan Adeli, and Dinggang Shen. 2018. Landmark-based deep multi-instance learning for brain disease diagnosis. Med. Image Anal. 43 (2018), 157–168.Google ScholarCross Ref
- Ningning Ma, Xiangyu Zhang, Hai-Tao Zheng, and Jian Sun. 2018. Shufflenet v2: Practical guidelines for efficient CNN architecture design. In Proceedings of the European Conference on Computer Vision. 116–131.Google ScholarDigital Library
- Tapas R. Martha, Norman Kerle, Victor Jetten, Cees J. van Westen, and K. Vinod Kumar. 2010. Characterising spectral, spatial and morphometric properties of landslides for semi-automatic detection using object-oriented methods. Geomorphology 116, 1–2 (2010), 24–36.Google ScholarCross Ref
- Vahid Moosavi, Ali Talebi, and Bagher Shirmohammadi. 2014. Producing a landslide inventory map using pixel-based and object-oriented approaches optimized by Taguchi method. Geomorphology 204 (2014), 646–656.Google ScholarCross Ref
- J. Nichol and M. S. Wong. 2005. Detection and interpretation of landslides using satellite images. Land Degrad. Devel. 16, 3 (2005), 243–255.Google ScholarCross Ref
- Arianna Pesci, Giordano Teza, Giuseppe Casula, Fabiana Loddo, Prospero De Martino, Mario Dolce, Francesco Obrizzo, and Folco Pingue. 2011. Multitemporal laser scanner-based observation of the Mt. Vesuvius crater: Characterization of overall geometry and recognition of landslide events. ISPRS J. Photogram. Remote Sens. 66, 3 (2011), 327–336.Google ScholarCross Ref
- Hamid Reza Pourghasemi and Norman Kerle. 2016. Random forests and evidential belief function-based landslide susceptibility assessment in Western Mazandaran Province, Iran. Environ. Earth Sci. 75, 3 (2016), 185.Google ScholarCross Ref
- Maoying Qiao, Liu Liu, Jun Yu, Chang Xu, and Dacheng Tao. 2017. Diversified dictionaries for multi-instance learning. Pattern Recog. 64 (2017), 407–416. Google ScholarDigital Library
- Hiroshi P. Sato, Hiroyuki Hasegawa, Satoshi Fujiwara, Mikio Tobita, Mamoru Koarai, Hiroshi Une, and Junko Iwahashi. 2007. Interpretation of landslide distribution triggered by the 2005 Northern Pakistan earthquake using SPOT 5 imagery. Landslides 4, 2 (2007), 113–122.Google ScholarCross Ref
- Shunping Ji, Dawen Yu, Chaoyong Shen, Weile Li, and Qiang Xu. 2020. Landslide detection from an open satellite imagery and digital elevation model dataset using attention boosted convolutional neural networks. Landslides 17, 6 (2020), 1337--1352.Google ScholarCross Ref
- Wenzhong Shi, Min Zhang, Hongfei Ke, Xin Fang, Zhao Zhan, and Shanxiong Chen. 2020. Landslide recognition by deep convolutional neural network and change detection. IEEE Trans. Geosci. Remote Sens. (21 Aug. 2020) (early access).Google ScholarCross Ref
- Karen Simonyan and Andrew Zisserman. 2015. Very deep convolutional networks for large-scale image recognition. In Proceedings of the International Conference on Learning Representations.Google Scholar
- Mingxing Tan and Quoc Le. 2019. Efficientnet: Rethinking model scaling for convolutional neural networks. In Proceedings of the International Conference on Machine Learning. 6105–6114.Google Scholar
- David J. Varnes. 1958. Landslide types and processes. Landsl. Eng. Pract. 24 (1958), 20–47.Google Scholar
- Ruili Wang, Wanting Ji, Mingzhe Liu, Xun Wang, Jian Weng, Song Deng, Suying Gao, and Chang-an Yuan. 2018. Review on mining data from multiple data sources. Pattern Recog. Lett. 109 (2018), 120–128.Google ScholarCross Ref
- Shujun Wang, Yaxi Zhu, Lequan Yu, Hao Chen, Huangjing Lin, Xiangbo Wan, Xinjuan Fan, and Pheng-Ann Heng. 2019. RMDL: Recalibrated multi-instance deep learning for whole slide gastric image classification. Med. Image Anal. 58 (2019), 101549.Google ScholarCross Ref
- Xiu-Shen Wei, Jianxin Wu, and Zhi-Hua Zhou. 2016. Scalable algorithms for multi-instance learning. IEEE Trans. Neural Netw. Learn. Syst. 28, 4 (2016), 975–987.Google ScholarCross Ref
- Gerald F. Wieczorek. 1984. Preparing a detailed landslide-inventory map for hazard evaluation and reduction. Bull. Assoc. Eng. Geol. 21, 3 (1984), 337–342.Google Scholar
- Jia Wu, Shirui Pan, Xingquan Zhu, Chengqi Zhang, and Xindong Wu. 2018. Multi-instance learning with discriminative bag mapping. IEEE Trans. Knowl. Data Eng. 30, 6 (2018), 1065–1080.Google ScholarCross Ref
- Chong Xu, Xiwei Xu, and J. Bruce H. Shyu. 2015. Database and spatial distribution of landslides triggered by the Lushan, China Mw 6.6 earthquake of 20 April 2013. Geomorphology 248 (2015), 77–92.Google Scholar
- Yaning Yi and Wanchang Zhang. 2020. A new deep-learning-based approach for earthquake-triggered landslide detection from single-temporal rapideye satellite imagery. IEEE J. Select. Topics Appl. Earth Observ. Remote Sens. 13 (2020), 6166–6176.Google ScholarCross Ref
- Wei Zhao, Ainong Li, Xi Nan, Zhengjian Zhang, and Guangbin Lei. 2017. Postearthquake landslides mapping from Landsat-8 data for the 2015 Nepal earthquake using a pixel-based change detection method. IEEE J. Select. Topics Appl. Earth Observ. Remote Sens. 10, 5 (2017), 1758–1768.Google ScholarCross Ref
- Yizhou Zhou, Xiaoyan Sun, Dong Liu, Zhengjun Zha, and Wenjun Zeng. 2017. Adaptive pooling in multi-instance learning for web video annotation. In Proceedings of the IEEE International Conference on Computer Vision Workshops. 318–327.Google Scholar
Index Terms
- MILL: Channel Attention–based Deep Multiple Instance Learning for Landslide Recognition
Recommendations
Multiple instance learning with bag dissimilarities
Multiple instance learning (MIL) is concerned with learning from sets (bags) of objects (instances), where the individual instance labels are ambiguous. In this setting, supervised learning cannot be applied directly. Often, specialized MIL methods ...
Multiple instance learning
The characteristics specific of MIL problems are formally identified and described.MIL methods and applications are reviewed in the light of the problem characteristics.Comparative experiments show the impact of problem characteristics on 16 reference ...
Revisiting multiple instance neural networks
We revisit the problem of solving MIL using neural networks (MINNs), which are ignored in current MIL research community. Our experiments show that MINNs are very effective and efficient.We proposed a novel MI-Net which is centered on learning bag ...
Comments