Self-attention and generative adversarial networks for algae monitoring
Water is important for the natural environment and human health. Monitoring algae concentrations yield information on the water quality. Compared with in situ measurements of water quality parameters, which are often complex and expensive, remote sensing techniques, using hyperspectral data analysis...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2022-12-01
|
| Series: | European Journal of Remote Sensing |
| Subjects: | |
| Online Access: | http://dx.doi.org/10.1080/22797254.2021.2010605 |
| _version_ | 1798034102719348736 |
|---|---|
| author | Nhut Hai Huynh Gordon Böer Hauke Schramm |
| author_facet | Nhut Hai Huynh Gordon Böer Hauke Schramm |
| author_sort | Nhut Hai Huynh |
| collection | DOAJ |
| description | Water is important for the natural environment and human health. Monitoring algae concentrations yield information on the water quality. Compared with in situ measurements of water quality parameters, which are often complex and expensive, remote sensing techniques, using hyperspectral data analysis, are fast and cost-effective. The objectives of this study are (1) to estimate the algae concentrations from hyperspectral data using deep learning techniques, (2) to investigate the applicability of attention mechanisms in the analysis of hyperspectral data, and (3) to augment the training data using generative adversarial networks (GANs). The results show that the accuracy of deep learning techniques is 7.6% higher than that of simpler artificial neural networks. Compared to noise injection and principal component analysis-based data augmentation, the use of a GAN-based data augmentation method significantly improves the accuracy of algae concentration estimates (>5%). In addition, models with added attention mechanisms yield an on average 3.13% higher accuracy than those without attention techniques. This result demonstrates the improvement of spectral features of artificial hyperspectral data based on the self-attention approach, revealing the potential of attention techniques in hyperspectral remote sensing. |
| first_indexed | 2024-04-11T20:39:44Z |
| format | Article |
| id | doaj.art-41608afd08cb41f8aca819f64255ebbb |
| institution | Directory Open Access Journal |
| issn | 2279-7254 |
| language | English |
| last_indexed | 2024-04-11T20:39:44Z |
| publishDate | 2022-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | European Journal of Remote Sensing |
| spelling | doaj.art-41608afd08cb41f8aca819f64255ebbb2022-12-22T04:04:16ZengTaylor & Francis GroupEuropean Journal of Remote Sensing2279-72542022-12-01551102210.1080/22797254.2021.20106052010605Self-attention and generative adversarial networks for algae monitoringNhut Hai Huynh0Gordon BöerHauke Schramm1Bbe Moldaenke GmbHKiel University of Applied SciencesWater is important for the natural environment and human health. Monitoring algae concentrations yield information on the water quality. Compared with in situ measurements of water quality parameters, which are often complex and expensive, remote sensing techniques, using hyperspectral data analysis, are fast and cost-effective. The objectives of this study are (1) to estimate the algae concentrations from hyperspectral data using deep learning techniques, (2) to investigate the applicability of attention mechanisms in the analysis of hyperspectral data, and (3) to augment the training data using generative adversarial networks (GANs). The results show that the accuracy of deep learning techniques is 7.6% higher than that of simpler artificial neural networks. Compared to noise injection and principal component analysis-based data augmentation, the use of a GAN-based data augmentation method significantly improves the accuracy of algae concentration estimates (>5%). In addition, models with added attention mechanisms yield an on average 3.13% higher accuracy than those without attention techniques. This result demonstrates the improvement of spectral features of artificial hyperspectral data based on the self-attention approach, revealing the potential of attention techniques in hyperspectral remote sensing.http://dx.doi.org/10.1080/22797254.2021.2010605deep learningself-attentiongenerative adversarial networkspcahyperspectral data augmentationremote sensing |
| spellingShingle | Nhut Hai Huynh Gordon Böer Hauke Schramm Self-attention and generative adversarial networks for algae monitoring European Journal of Remote Sensing deep learning self-attention generative adversarial networks pca hyperspectral data augmentation remote sensing |
| title | Self-attention and generative adversarial networks for algae monitoring |
| title_full | Self-attention and generative adversarial networks for algae monitoring |
| title_fullStr | Self-attention and generative adversarial networks for algae monitoring |
| title_full_unstemmed | Self-attention and generative adversarial networks for algae monitoring |
| title_short | Self-attention and generative adversarial networks for algae monitoring |
| title_sort | self attention and generative adversarial networks for algae monitoring |
| topic | deep learning self-attention generative adversarial networks pca hyperspectral data augmentation remote sensing |
| url | http://dx.doi.org/10.1080/22797254.2021.2010605 |
| work_keys_str_mv | AT nhuthaihuynh selfattentionandgenerativeadversarialnetworksforalgaemonitoring AT gordonboer selfattentionandgenerativeadversarialnetworksforalgaemonitoring AT haukeschramm selfattentionandgenerativeadversarialnetworksforalgaemonitoring |