Taxonomic Identification of European Seahorse H. Guttulatus and H. Hippocampus Based on Machine Learning Techniques
In recent years, machine learning (ML) algorithms have emerged as a powerful and innovative tool to enhance data processing and facilitate species identification across taxa, including birds, insects, and plants. Transfer learning can be used for classification, regression and clustering problems. This paper uses two trained models - VGG16 and ResNet50 - with Deep Convolutional Neural Networks for image recognition. We apply this methodology to Seahorse species recognition. The experimental results show that the average accuracy of VGG16 network can reach 91.7%.
Authors
- Simone Marini (ORCID: https://orcid.org/0000-0002-5704-3533)
- Marco Spoto
- Daniele Lagomarsino-Oneto (ORCID: https://orcid.org/0000-0002-9842-5570)
- Michele Gristina (ORCID: https://orcid.org/0000-0003-3639-7655)
- Alessandro Rinaldi (ORCID: https://orcid.org/0000-0002-5237-4990)
- Cataldo Pierri (ORCID: https://orcid.org/0000-0002-6125-3322)
- Mattia Cavaiola (ORCID: https://orcid.org/0000-0002-1316-9484)
Institutions
- Istituto di Scienze Marine del Consiglio Nazionale delle Ricerche (IT)
- Institute of Structure of Matter (IT)
- University of Bari Aldo Moro (IT)
- University of Bologna (IT)
Publication Details
- Published
- 2024-10-14
- DOI
- https://doi.org/10.1109/metrosea62823.2024.10765623
- Primary Topic
- Aquatic life and conservation
- Type
- article
- Field-Weighted Citation Impact
- 0.00