The SARFish Dataset and Challenge

Connor Luckett, Benjamin McCarthy, Tri-Tan Cao, Antonio Robles-Kelly; Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops, 2024, pp. 752-761

Abstract


In this paper, we present the SARFish challenge and dataset. The challenge focuses on the use of using synthetic aperture radar (SAR) imagery for the identification of vessels involved in illegal and unregulated fishing. Illegal, unreported, and unregulated fishing damages ecological systems and causes losses for fishing industries and governments worldwide. The SARFish dataset is a free and open large-scale complex-valued SAR dataset which is based upon Sentinel-1 imagery and builds upon xView3. We expect this dataset to advance the state of the art in automated detection from SAR imagery, contextual representation learning and deep complex-valued networks. We also hope the availability of the SARFish dataset will stimulate developments on other topics of interest that can naturally tackle complex-valued data such as quantum-inspired approaches.

Related Material


[pdf]
[bibtex]
@InProceedings{Luckett_2024_WACV, author = {Luckett, Connor and McCarthy, Benjamin and Cao, Tri-Tan and Robles-Kelly, Antonio}, title = {The SARFish Dataset and Challenge}, booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops}, month = {January}, year = {2024}, pages = {752-761} }