SFISHTRACK: A Dataset for Fish Segmentation and Tracking in Underwater Videos

Read full text
Authors: Josep S. Sánchez, José-Luis Lisani, Ignacio A. Catalán, Amaya Álvarez-Ellacuría
Cited by: 1

Abstract

A high-quality underwater video dataset for fish segmentation and multi-object tracking, featuring pixel-level annotations, temporally consistent identities, and structured metadata from diverse marine environments in the Balearic Sea. Designed for benchmarking and developing underwater computer vision methods. To refer to scripts used for visualizing the data please consult the github link added to the code repository found in this page. In case of need for citation we encourage to follow the instructions written at the end of the Github page. For further questions please contact Josep, Sánchez.

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
1 June 2026
Language
EN
Document type
Dataset