Lake Ladoga in Russia and Lake Saimaa in Finland are inhabited by endangered endemic ringed seal population. Nowadays, the major threats for both seals are related to human activities and exaggerated by growing effects of climate change. Interactions between nature conservation needs and local socio-economic interests are complex and controversial, and form various trade-offs, which are difficult to solve on local level. For instance, the conflict between seals and fisheries in both lakes have existed for decades.

The main objectives of the project are 1) to ensure and strengthen the Russian-Finnish cross-border cooperation in the field of conservation biology, 2) to take coordinated science based actions towards human-freshwater seal co-existence via conservation biology and novel technologies enabling automatic cost-effective monitoring of seals, and 3) to enhance environmental awareness among local residents, visitors and companies on both sides of the border.

Compuater Vision and Pattern Recognition Laboratory (CVPRL) at LUT University is responsible for developing novel seal monitoring tools via computer vision. Game camera traps produce typically massive amount of image material. Monitoring requires identifying of individual seals from the images which if done by human experts is very time-consuming. Therefore, there is an urgent need for automated methods for the identification. Ringed seals have a distinctive permanent pelage pattern that is unique to each individual making the image-based identification possible. The main approach for developing identification methods will be deep learning and convolutional neural networks that have been shown to produce state-of-the-art accuracy for various computer vision tasks. The developed method will be evaluated and analyzed thoroughly using the data collected during the project. The identification method development continues the work that was done in the SealVision project, where the indvidual identification of Saimaa ringed seals was studied.

The CoExist is funded from the South-East Finland - Russia CBC 2014-2020 programme. It is a collaborative project between LUT, University of Eastern Finland (lead partner), Interregional charitable public organisation Biologists for nature conservation from Russia, and The Finnish Association for Nature Conservation, South Karelia division.

South-Eastern Finland-Russia CBC 2014-2020
University of Eastern Finland    LUT University    Biologists for nature conservation    Finnish Association for Nature Conservation


Ekaterina Nepovinnykh E-mail Office: 2419 Doctoral student
Tuomas Eerola E-mail Office: 2425 Tel: +358 50 404 2054 Associate Professor, Project Manager
Heikki Kälviäinen E-mail Office: 2422 Tel: +358 40 586 7552 Professor, Project Leader


EDEN: Deep feature distribution pooling for saimaa ringed seals pattern matching

By Ilja Chelak, Ekaterina Nepovinnykh, Tuomas Eerola, Heikki Kalviainen, and Igor Belykh, Accepted to International Conference on Cyber-Physical Systems & Control, 2021.

Download: PDF arXiv

Metric learning based pattern matching for species agnostic animal re-identification

By Ola Badreldeen Bdawy Mohamed, Master's Thesis, LUT University, 2021.

Download: URN

Seal pose estimation using convolutional neural networks

By Yordanos Alemu, Master's Thesis, LUT University, 2021.

Download: URN

Siamese Network Based Pelage Pattern Matching for Ringed Seal Re-identification

By Ekaterina Nepovinnykh, Tuomas Eerola, and Heikki Kälviäinen, In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops, 2020.

Download: PDF DOI

Instance segmentation of Ladoga ringed seals

By Andrei Lushpanov, Master's Thesis, LUT University, 2020.

Download: PDF URN

CNN-based ringed seal pelage pattern extraction

By Denis Zavialkin, Master's Thesis, LUT University, 2020.

Download: PDF URN



CoExist project main website