Towards Visual Re-Identification of Fish using Fine-Grained Classification for Electronic Monitoring in Fisheries

Mahagedara Waththe Samitha Nuwan Thilakarathna, Ercan Avsar, Martin Mathias Nielsen, Malte Pedersen
Proceedings of the 7th Northern Lights Deep Learning Conference (NLDL), PMLR 307:428-438, 2026.

Abstract

Accurate fisheries data are crucial for effective and sustainable marine resource management. With the recent adoption of Electronic Monitoring (EM) systems, more video data is now being collected than can be feasibly reviewed manually. This paper addresses this challenge by developing an optimized deep learning pipeline for automated fish re-identification (Re-ID) using the novel AutoFish dataset, which simulates EM systems with conveyor belts with six similarly looking fish species. We demonstrate that key Re-ID metrics (R1 and mAP@k) are substantially improved by using hard triplet mining in conjunction with a custom image transformation pipeline that includes dataset-specific normalization. By employing these strategies, we demonstrate that the Vision Transformer-based Swin-T architecture consistently outperforms the Convolutional Neural Network-based ResNet-50, achieving peak performance of 41.65% mAP@k and 90.43% Rank-1 accuracy. An in-depth analysis reveals that the primary challenge is distinguishing visually similar individuals of the same species (Intra-species errors), where viewpoint inconsistency proves significantly more detrimental than partial occlusion. The source code and documentation are available at: \url{https://github.com/msamdk/Fish_Re_Identification.git}

Cite this Paper

BibTeX
@InProceedings{pmlr-v307-thilakarathna26a, title = {Towards Visual Re-Identification of Fish using Fine-Grained Classification for Electronic Monitoring in Fisheries}, author = {Thilakarathna, Mahagedara Waththe Samitha Nuwan and Avsar, Ercan and Nielsen, Martin Mathias and Pedersen, Malte}, booktitle = {Proceedings of the 7th Northern Lights Deep Learning Conference (NLDL)}, pages = {428--438}, year = {2026}, editor = {Kim, Hyeongji and Ramírez Rivera, Adín and Ricaud, Benjamin}, volume = {307}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jan}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v307/main/assets/thilakarathna26a/thilakarathna26a.pdf}, url = {https://proceedings.mlr.press/v307/thilakarathna26a.html}, abstract = {Accurate fisheries data are crucial for effective and sustainable marine resource management. With the recent adoption of Electronic Monitoring (EM) systems, more video data is now being collected than can be feasibly reviewed manually. This paper addresses this challenge by developing an optimized deep learning pipeline for automated fish re-identification (Re-ID) using the novel AutoFish dataset, which simulates EM systems with conveyor belts with six similarly looking fish species. We demonstrate that key Re-ID metrics (R1 and mAP@k) are substantially improved by using hard triplet mining in conjunction with a custom image transformation pipeline that includes dataset-specific normalization. By employing these strategies, we demonstrate that the Vision Transformer-based Swin-T architecture consistently outperforms the Convolutional Neural Network-based ResNet-50, achieving peak performance of 41.65% mAP@k and 90.43% Rank-1 accuracy. An in-depth analysis reveals that the primary challenge is distinguishing visually similar individuals of the same species (Intra-species errors), where viewpoint inconsistency proves significantly more detrimental than partial occlusion. The source code and documentation are available at: \url{https://github.com/msamdk/Fish_Re_Identification.git}} }
Endnote
%0 Conference Paper %T Towards Visual Re-Identification of Fish using Fine-Grained Classification for Electronic Monitoring in Fisheries %A Mahagedara Waththe Samitha Nuwan Thilakarathna %A Ercan Avsar %A Martin Mathias Nielsen %A Malte Pedersen %B Proceedings of the 7th Northern Lights Deep Learning Conference (NLDL) %C Proceedings of Machine Learning Research %D 2026 %E Hyeongji Kim %E Adín Ramírez Rivera %E Benjamin Ricaud %F pmlr-v307-thilakarathna26a %I PMLR %P 428--438 %U https://proceedings.mlr.press/v307/thilakarathna26a.html %V 307 %X Accurate fisheries data are crucial for effective and sustainable marine resource management. With the recent adoption of Electronic Monitoring (EM) systems, more video data is now being collected than can be feasibly reviewed manually. This paper addresses this challenge by developing an optimized deep learning pipeline for automated fish re-identification (Re-ID) using the novel AutoFish dataset, which simulates EM systems with conveyor belts with six similarly looking fish species. We demonstrate that key Re-ID metrics (R1 and mAP@k) are substantially improved by using hard triplet mining in conjunction with a custom image transformation pipeline that includes dataset-specific normalization. By employing these strategies, we demonstrate that the Vision Transformer-based Swin-T architecture consistently outperforms the Convolutional Neural Network-based ResNet-50, achieving peak performance of 41.65% mAP@k and 90.43% Rank-1 accuracy. An in-depth analysis reveals that the primary challenge is distinguishing visually similar individuals of the same species (Intra-species errors), where viewpoint inconsistency proves significantly more detrimental than partial occlusion. The source code and documentation are available at: \url{https://github.com/msamdk/Fish_Re_Identification.git}
APA
Thilakarathna, M.W.S.N., Avsar, E., Nielsen, M.M. & Pedersen, M.. (2026). Towards Visual Re-Identification of Fish using Fine-Grained Classification for Electronic Monitoring in Fisheries. Proceedings of the 7th Northern Lights Deep Learning Conference (NLDL), in Proceedings of Machine Learning Research 307:428-438 Available from https://proceedings.mlr.press/v307/thilakarathna26a.html.

Related Material