Deep Learning Classification of Lake Zooplankton
Plankton are effective indicators of environmental change and ecosystem health in freshwater habitats, but collection of plankton data using manual microscopic methods is extremely labor-intensive and expensive. Automated plankton imaging offers a promising way forward to monitor plankton communitie...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2021-11-01
|
| Series: | Frontiers in Microbiology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmicb.2021.746297/full |
| _version_ | 1818838506648109056 |
|---|---|
| author | Sreenath P. Kyathanahally Thomas Hardeman Ewa Merz Thea Bulas Marta Reyes Peter Isles Francesco Pomati Marco Baity-Jesi |
| author_facet | Sreenath P. Kyathanahally Thomas Hardeman Ewa Merz Thea Bulas Marta Reyes Peter Isles Francesco Pomati Marco Baity-Jesi |
| author_sort | Sreenath P. Kyathanahally |
| collection | DOAJ |
| description | Plankton are effective indicators of environmental change and ecosystem health in freshwater habitats, but collection of plankton data using manual microscopic methods is extremely labor-intensive and expensive. Automated plankton imaging offers a promising way forward to monitor plankton communities with high frequency and accuracy in real-time. Yet, manual annotation of millions of images proposes a serious challenge to taxonomists. Deep learning classifiers have been successfully applied in various fields and provided encouraging results when used to categorize marine plankton images. Here, we present a set of deep learning models developed for the identification of lake plankton, and study several strategies to obtain optimal performances, which lead to operational prescriptions for users. To this aim, we annotated into 35 classes over 17900 images of zooplankton and large phytoplankton colonies, detected in Lake Greifensee (Switzerland) with the Dual Scripps Plankton Camera. Our best models were based on transfer learning and ensembling, which classified plankton images with 98% accuracy and 93% F1 score. When tested on freely available plankton datasets produced by other automated imaging tools (ZooScan, Imaging FlowCytobot, and ISIIS), our models performed better than previously used models. Our annotated data, code and classification models are freely available online. |
| first_indexed | 2024-12-19T03:39:29Z |
| format | Article |
| id | doaj.art-3e20fa8149824e41b8e6f4721ec9e236 |
| institution | Directory Open Access Journal |
| issn | 1664-302X |
| language | English |
| last_indexed | 2024-12-19T03:39:29Z |
| publishDate | 2021-11-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Microbiology |
| spelling | doaj.art-3e20fa8149824e41b8e6f4721ec9e2362022-12-21T20:37:17ZengFrontiers Media S.A.Frontiers in Microbiology1664-302X2021-11-011210.3389/fmicb.2021.746297746297Deep Learning Classification of Lake ZooplanktonSreenath P. KyathanahallyThomas HardemanEwa MerzThea BulasMarta ReyesPeter IslesFrancesco PomatiMarco Baity-JesiPlankton are effective indicators of environmental change and ecosystem health in freshwater habitats, but collection of plankton data using manual microscopic methods is extremely labor-intensive and expensive. Automated plankton imaging offers a promising way forward to monitor plankton communities with high frequency and accuracy in real-time. Yet, manual annotation of millions of images proposes a serious challenge to taxonomists. Deep learning classifiers have been successfully applied in various fields and provided encouraging results when used to categorize marine plankton images. Here, we present a set of deep learning models developed for the identification of lake plankton, and study several strategies to obtain optimal performances, which lead to operational prescriptions for users. To this aim, we annotated into 35 classes over 17900 images of zooplankton and large phytoplankton colonies, detected in Lake Greifensee (Switzerland) with the Dual Scripps Plankton Camera. Our best models were based on transfer learning and ensembling, which classified plankton images with 98% accuracy and 93% F1 score. When tested on freely available plankton datasets produced by other automated imaging tools (ZooScan, Imaging FlowCytobot, and ISIIS), our models performed better than previously used models. Our annotated data, code and classification models are freely available online.https://www.frontiersin.org/articles/10.3389/fmicb.2021.746297/fullplankton cameradeep learningplankton classificationtransfer learningGreifenseeensemble learning |
| spellingShingle | Sreenath P. Kyathanahally Thomas Hardeman Ewa Merz Thea Bulas Marta Reyes Peter Isles Francesco Pomati Marco Baity-Jesi Deep Learning Classification of Lake Zooplankton Frontiers in Microbiology plankton camera deep learning plankton classification transfer learning Greifensee ensemble learning |
| title | Deep Learning Classification of Lake Zooplankton |
| title_full | Deep Learning Classification of Lake Zooplankton |
| title_fullStr | Deep Learning Classification of Lake Zooplankton |
| title_full_unstemmed | Deep Learning Classification of Lake Zooplankton |
| title_short | Deep Learning Classification of Lake Zooplankton |
| title_sort | deep learning classification of lake zooplankton |
| topic | plankton camera deep learning plankton classification transfer learning Greifensee ensemble learning |
| url | https://www.frontiersin.org/articles/10.3389/fmicb.2021.746297/full |
| work_keys_str_mv | AT sreenathpkyathanahally deeplearningclassificationoflakezooplankton AT thomashardeman deeplearningclassificationoflakezooplankton AT ewamerz deeplearningclassificationoflakezooplankton AT theabulas deeplearningclassificationoflakezooplankton AT martareyes deeplearningclassificationoflakezooplankton AT peterisles deeplearningclassificationoflakezooplankton AT francescopomati deeplearningclassificationoflakezooplankton AT marcobaityjesi deeplearningclassificationoflakezooplankton |