Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species
Automated image analysis and deep learning tools such as object detection models are being used increasingly by biologists. However, biological datasets often have constraints that are challenging for the use of deep learning. Classes are often imbalanced, similar, or too few for robust learning. In...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9411844/ |
| _version_ | 1818850474776854528 |
|---|---|
| author | Paul Tresson Dominique Carval Philippe Tixier William Puech |
| author_facet | Paul Tresson Dominique Carval Philippe Tixier William Puech |
| author_sort | Paul Tresson |
| collection | DOAJ |
| description | Automated image analysis and deep learning tools such as object detection models are being used increasingly by biologists. However, biological datasets often have constraints that are challenging for the use of deep learning. Classes are often imbalanced, similar, or too few for robust learning. In this paper we present a robust method relying on hierarchical classification to perform very small object detection. We illustrate our results on a custom dataset featuring 22 classes of arthropods used to study biodiversity. This dataset shows several constraints that are frequent when using deep learning on biological data with a high class imbalance, some classes learned on only a few training examples and a high similarity between classes. We propose to first perform detection at a super-class level, before performing a detailed classification at a class level. We compare the obtained results with our proposed method to a global detector, trained without hierarchical classification. Our method succeeds in obtaining a mAP of 75 %, while the global detector only achieves a mAP of 48 %. Moreover, our method shows high precision even on classes with the less train examples. Confusions between classes with our method are fewer and are of a lesser impact. We are able achieve a more robust object classification with the use of our proposed method. This method can also enable better control on the model’s output which can be particularly valuable when handling ecological, biological or medical data for example. |
| first_indexed | 2024-12-19T06:49:43Z |
| format | Article |
| id | doaj.art-550d4f06fd6e4949adf5d4a4df9782ae |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-19T06:49:43Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-550d4f06fd6e4949adf5d4a4df9782ae2022-12-21T20:31:47ZengIEEEIEEE Access2169-35362021-01-019639256393210.1109/ACCESS.2021.30752939411844Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod SpeciesPaul Tresson0https://orcid.org/0000-0002-1275-4673Dominique Carval1https://orcid.org/0000-0002-1525-2855Philippe Tixier2https://orcid.org/0000-0001-5147-9777William Puech3https://orcid.org/0000-0001-9383-2401CIRAD, UPR GECO, Montpellier, FranceCIRAD, UPR GECO, Montpellier, FranceCIRAD, UPR GECO, Montpellier, FranceLIRMM, Univ. Montpellier, CNRS, Montpellier, FranceAutomated image analysis and deep learning tools such as object detection models are being used increasingly by biologists. However, biological datasets often have constraints that are challenging for the use of deep learning. Classes are often imbalanced, similar, or too few for robust learning. In this paper we present a robust method relying on hierarchical classification to perform very small object detection. We illustrate our results on a custom dataset featuring 22 classes of arthropods used to study biodiversity. This dataset shows several constraints that are frequent when using deep learning on biological data with a high class imbalance, some classes learned on only a few training examples and a high similarity between classes. We propose to first perform detection at a super-class level, before performing a detailed classification at a class level. We compare the obtained results with our proposed method to a global detector, trained without hierarchical classification. Our method succeeds in obtaining a mAP of 75 %, while the global detector only achieves a mAP of 48 %. Moreover, our method shows high precision even on classes with the less train examples. Confusions between classes with our method are fewer and are of a lesser impact. We are able achieve a more robust object classification with the use of our proposed method. This method can also enable better control on the model’s output which can be particularly valuable when handling ecological, biological or medical data for example.https://ieeexplore.ieee.org/document/9411844/Very small object detectiondeep learningrobustnessbiodiversitytaxonomy |
| spellingShingle | Paul Tresson Dominique Carval Philippe Tixier William Puech Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species IEEE Access Very small object detection deep learning robustness biodiversity taxonomy |
| title | Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species |
| title_full | Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species |
| title_fullStr | Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species |
| title_full_unstemmed | Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species |
| title_short | Hierarchical Classification of Very Small Objects: Application to the Detection of Arthropod Species |
| title_sort | hierarchical classification of very small objects application to the detection of arthropod species |
| topic | Very small object detection deep learning robustness biodiversity taxonomy |
| url | https://ieeexplore.ieee.org/document/9411844/ |
| work_keys_str_mv | AT paultresson hierarchicalclassificationofverysmallobjectsapplicationtothedetectionofarthropodspecies AT dominiquecarval hierarchicalclassificationofverysmallobjectsapplicationtothedetectionofarthropodspecies AT philippetixier hierarchicalclassificationofverysmallobjectsapplicationtothedetectionofarthropodspecies AT williampuech hierarchicalclassificationofverysmallobjectsapplicationtothedetectionofarthropodspecies |