Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey
In deep learning-based hyperspectral remote sensing image classification tasks, random sampling strategies are typically used to train model parameters for testing and evaluation. However, this approach leads to strong spatial autocorrelation between the training set samples and the surrounding test...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/15/3793 |
| _version_ | 1827730919503429632 |
|---|---|
| author | Hao Feng Yongcheng Wang Zheng Li Ning Zhang Yuxi Zhang Yunxiao Gao |
| author_facet | Hao Feng Yongcheng Wang Zheng Li Ning Zhang Yuxi Zhang Yunxiao Gao |
| author_sort | Hao Feng |
| collection | DOAJ |
| description | In deep learning-based hyperspectral remote sensing image classification tasks, random sampling strategies are typically used to train model parameters for testing and evaluation. However, this approach leads to strong spatial autocorrelation between the training set samples and the surrounding test set samples, and some unlabeled test set data directly participate in the training of the network. This leaked information makes the model overly optimistic. Models trained under these conditions tend to overfit to a single dataset, which limits the range of practical applications. This paper analyzes the causes and effects of information leakage and summarizes the methods from existing models to mitigate the effects of information leakage. Specifically, this paper states the main issues in this area, where the issue of information leakage is addressed in detail. Second, some algorithms and related models used to mitigate information leakage are categorized, including reducing the number of training samples, using spatially disjoint sampling strategies, few-shot learning, and unsupervised learning. These models and methods are classified according to the sample-related phase and the feature extraction phase. Finally, several representative hyperspectral image classification models experiments are conducted on the common datasets and their effectiveness in mitigating information leakage is analyzed. |
| first_indexed | 2024-03-11T00:17:23Z |
| format | Article |
| id | doaj.art-50571a42fd6d48598c9bb34567907552 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T00:17:23Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-50571a42fd6d48598c9bb345679075522023-11-18T23:30:52ZengMDPI AGRemote Sensing2072-42922023-07-011515379310.3390/rs15153793Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A SurveyHao Feng0Yongcheng Wang1Zheng Li2Ning Zhang3Yuxi Zhang4Yunxiao Gao5Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, ChinaDepartment of Electronic Engineering, Tsinghua University, Beijing 100084, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, ChinaIn deep learning-based hyperspectral remote sensing image classification tasks, random sampling strategies are typically used to train model parameters for testing and evaluation. However, this approach leads to strong spatial autocorrelation between the training set samples and the surrounding test set samples, and some unlabeled test set data directly participate in the training of the network. This leaked information makes the model overly optimistic. Models trained under these conditions tend to overfit to a single dataset, which limits the range of practical applications. This paper analyzes the causes and effects of information leakage and summarizes the methods from existing models to mitigate the effects of information leakage. Specifically, this paper states the main issues in this area, where the issue of information leakage is addressed in detail. Second, some algorithms and related models used to mitigate information leakage are categorized, including reducing the number of training samples, using spatially disjoint sampling strategies, few-shot learning, and unsupervised learning. These models and methods are classified according to the sample-related phase and the feature extraction phase. Finally, several representative hyperspectral image classification models experiments are conducted on the common datasets and their effectiveness in mitigating information leakage is analyzed.https://www.mdpi.com/2072-4292/15/15/3793hyperspectral image classificationspatial autocorrelationinformation leakagesampling strategy |
| spellingShingle | Hao Feng Yongcheng Wang Zheng Li Ning Zhang Yuxi Zhang Yunxiao Gao Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey Remote Sensing hyperspectral image classification spatial autocorrelation information leakage sampling strategy |
| title | Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey |
| title_full | Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey |
| title_fullStr | Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey |
| title_full_unstemmed | Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey |
| title_short | Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey |
| title_sort | information leakage in deep learning based hyperspectral image classification a survey |
| topic | hyperspectral image classification spatial autocorrelation information leakage sampling strategy |
| url | https://www.mdpi.com/2072-4292/15/15/3793 |
| work_keys_str_mv | AT haofeng informationleakageindeeplearningbasedhyperspectralimageclassificationasurvey AT yongchengwang informationleakageindeeplearningbasedhyperspectralimageclassificationasurvey AT zhengli informationleakageindeeplearningbasedhyperspectralimageclassificationasurvey AT ningzhang informationleakageindeeplearningbasedhyperspectralimageclassificationasurvey AT yuxizhang informationleakageindeeplearningbasedhyperspectralimageclassificationasurvey AT yunxiaogao informationleakageindeeplearningbasedhyperspectralimageclassificationasurvey |