Ringed seal image dataset (SealID) provides a challenging identification task. General purpose animal re-identification methods that utilize fur, feather, skin patterns of animals could be evaluated on that dataset. Large variation in illumination, seal poses, limited size of identifiable region, low contrast between the ring pattern and the rest of the pelage, substantial differences between wet and dry fur and low image quality of automatic camera traps all contribute to the difficulty of the re-identification task of Saimaa ringed seals. We have compiled an extensive dataset of 57 seals containing a total of 2080 images and made it publicly available. All animal individuals were identified by experts. If you use this database in your research, we kindly ask that you reference our papers listed below: Nepovinnykh, E., Eerola, T., Biard, V., Mutka, P., Niemi, M., Kunnasranta, M. and Kälviäinen, H., 2022. SealID: Saimaa ringed seal re-identification dataset. Sensors, 22(19), p.7602. Nepovinnykh, E., Chelak, I., Eerola, T. and Kälviäinen, H., 2022. Norppa: Novel ringed seal re-identification by pelage pattern aggregation. arXiv preprint arXiv:2206.02498.

2022