Abstract
Computer vision can transform wildlife monitoring by automating phenotyping and individual identification. Achieving this, however, depends on access to large, well-curated image datasets that capture natural variation across individuals and years. Here, we present Melops, a longitudinal dataset comprising 24 578 images of 9 861 individual corkwing wrasse, Symphodus melops, collected over seven years through a capture–mark–recapture survey. Each fish was PIT-tagged for re-identification and photographed from both sides against a standardized white background with a colour reference. Alongside the images, we provide metadata including body length, sex, and reproductive state. To support deep learning applications, the dataset includes both the original photographs and automatically cropped images focusing on the whole fish or specific body regions. Together, these resources provide a foundation for developing computer vision methods for individual re-identification, colour pattern analysis, sex classification and other visual phenotyping tasks. Beyond this species, Melops can serve as a model for similar datasets in other taxa. Because it contains thousands of individuals with repeated observations, it provides a rare opportunity to explore temporally aware re-identification and phenotypic change in wild fish.
Similar content being viewed by others
Data availability
The current version of the dataset is available at Zenodo (https://doi.org/10.5281/zenodo.17404087). A summary of folders and files is provided in the readme.pdf. Additional data might be collected and uploaded in the future; in that case you can always find the most recent version of the dataset at Zenodo (https://doi.org/10.5281/zenodo.17099924).
Code availability
The code is available at https://doi.org/10.5281/zenodo.17404087. A summary of folders and files is provided in the readme.pdf.
References
Schneider, S., Taylor, G. W. & Kremer, S. C. Similarity learning networks for animal individual re-identification: an ecological perspective. Mammalian Biology 102, 899–914 (2022).
Goodwin, M. et al. Unlocking the potential of deep learning for marine ecology: Overview, applications, and outlook. ICES Journal of Marine Science 79, 319–336 (2022).
Weinstein, B. G. A computer vision for animal ecology. Journal of Animal Ecology 87, 533–545 (2017).
Schneider, S., Taylor, G. W., Linquist, S. & Kremer, S. C. Similarity Learning Networks for Animal Individual Re-Identification–Beyond the Capabilities of a Human Observer. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision 44–52 (2020).
Gómez-Vargas, N., Alonso-Fernández, A., Blanquero, R. & Antelo, L. T. Re-identification of fish individuals of undulate skate via deep learning within a few-shot context. Ecological Informatics 75, 102036 (2023).
Price, A. C., Weadick, C. J., Shim, J. & Rodd, F. H. Pigments, patterns, and fish behavior. Zebrafish 5, 297–307 (2008).
Adam, L., Papafitsoros, K., Kovář, R., Čermák, V. & Picek, L. Overview of AnimalCLEF 2025: Recognizing Individual Animals in Images. in Working Notes of CLEF 2025 — Conference and Labs of the Evaluation Forum, CEUR Workshop Proceedings, Vol. 4038 — (2025).
Wahltinez, O. & Wahltinez, S. J. An open- source general purpose machine learning framework for individual animal re- identification using few- shot learning. 2024, 1–15 (2024).
Olsen, Ø. L. et al. A contrastive learning approach for individual re-identification in a wild fish population. Proceedings of the Northern Lights Deep Learning Workshop 4 (2023).
Lonati, M. et al. Novel use of deep neural networks on photographic identification of epaulette sharks (Hemiscyllium ocellatum) across life stages. Journal of Fish Biology 105, 1572–1587 (2024).
Rosenberg, B. et al. Individual identification of brown bears using pose-aware metric learning. Current Biology 36, 645–659.e14 (2026).
Li, Y. et al. MetaWild: A Multimodal Dataset for Animal Re-Identification with Environmental Metadata. in Proceedings of the 33rd ACM International Conference on Multimedia 13009–13015, https://doi.org/10.1145/3746027.3758249 (ACM, Dublin Ireland, 2025).
Uglem, I., Rosenqvist, G. & Wasslavik, H. S. Phenotypic variation between dimorphic males in corkwing wrasse. Journal of Fish Biology 57, 1–14 (2000).
Halvorsen, K. T. et al. Male-biased sexual size dimorphism in the nest building corkwing wrasse (Symphodus melops): implications for a size regulated fishery. ICES Journal of Marine Science: Journal du Conseil 73, 2586–2594 (2016).
Alonzo, S. H. & Heckman, K. L. The unexpected but understandable dynamics of mating, paternity and paternal care in the ocellated wrasse. Proceedings. Biological sciences/The Royal Society 277, 115–22 (2010).
Taborsky, M., Hudde, B. & Wirtz, P. Reproductive behaviour and ecology of Symphodus (Crenilabrus) ocellatus, a European wrasse with four types of male behaviour. Behaviour http://www.jstor.org/stable/4534614 (1987).
Warner, R. R. & Lejeune, P. Sex change limited by paternal care: a test using four Mediterranean labrid fishes, genus Symphodus. Marine Biology 99, 89–99 (1985).
Halvorsen, K. T. et al. Movement patterns of temperate wrasses (Labridae) within a small marine protected area. Journal of Fish Biology jfb.14825, https://doi.org/10.1111/jfb.14825 (2021).
CVAT. ai Corporation. Computer Vision Annotation Tool (CVAT). Zenodo https://doi.org/10.5281/ZENODO.3497105 (2026).
Jocher, G., Chaurasia, A. & Jing, Q. Yolov8. (2023).
Ooms, J. Magick: Advanced Graphics and Image-Processing in R. (2025).
John, L., Rick, I. P., Vitt, S. & Thünken, T. Body coloration as a dynamic signal during intrasexual communication in a cichlid fish. BMC Zool 6, 9 (2021).
Gümüş, B., Gümüş, E. & Balaban, M. O. Color of rainbow trout (Oncorhynchus mykiss) fillets by image and sensory analysis, and correlation with SalmoFan numbers. Journal of Food Science 88, 430–446 (2023).
Van Der Walt, S. et al. scikit-image: image processing in Python. PeerJ 2, e453 (2014).
Sørdalen, T. K. et al. Melops: A wild fish image dataset for individual re-identification and phenotyping. Zenodo https://doi.org/10.5281/ZENODO.17404087 (2025).
Sauvaitre, C. A quantitative assessment of factors influencing sexual differences in colouration of Corkwing Wrasse (Symphodus melops). (University of Bordeaux, 2025).
Holmberg, J., Norman, B. & Arzoumanian, Z. Estimating population size, structure, and residency time for whale sharks Rhincodon typus through collaborative photo-identification. Endang. Species. Res. 7, 39–53 (2009).
Hitt, N. P., Kessler, K. G. & Letcher, B. H. Brook trout imagery data for individual recognition with deep learning: U.S. Geological Survey data release. https://doi.org/10.5066/P94UL1Z1.
Town, C., Marshall, A. & Sethasathien, N. Manta Matcher: automated photographic identification of manta rays using keypoint features. Ecology and Evolution 3, 1902–1914 (2013).
Pedersen, M. & Mohammed, A. Photo Identification of Individual Salmo trutta Based on Deep Learning. Applied Sciences 11, 9039 (2021).
McCoy, E. et al. Long-Term Photo-Identification Reveals the Population Dynamics and Strong Site Fidelity of Adult Whale Sharks to the Coastal Waters of Donsol, Philippines. Front. Mar. Sci. 5, 271 (2018).
Acknowledgements
We would like to thank our colleagues Anil Osman Tur and Vaneeda Allken for contributing to the FishFaces human benchmark test. We are also grateful to colleagues, students and interns who have assisted us in the field over the years. This work is funded by the Norwegian Research Council through the project “Computer Vision to Expand Monitoring and Accelerate Assessment of Coastal Fish (CoastVision)” with project number 325862 as well as the Norwegian Institute of Marine Research, project number 15638.
Funding
Open access funding provided by Institute Of Marine Research.
Author information
Authors and Affiliations
Contributions
T.K.S. and K.T.H. conceived, conceptualized and planned the study. K.T.H., T.L. and A.B.S. conducted the fieldwork. T.K.S. and K.T.H. annotated images, curated the dataset and investigated methodology. T.K.S. and K.T.H. constructed the datasets, analysed and visualized the data. C.S. annotated keypoints, ran the YOLOv8-pose model and evaluated keypoint detections. K.M. ran the YOLOv8x model. T.K.S. and K.T.H. wrote the original draft of the manuscript, with review and editing. C.S., K.M., C.B., A.B.S. and T.L. reviewed the manuscript. All authors approved the final version.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Sørdalen, T.K., Malde, K., Sauvaitre, C. et al. A wild fish image dataset for individual re-identification and phenotyping. Sci Data (2026). https://doi.org/10.1038/s41597-026-07045-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41597-026-07045-1
