This is version 1.0 of the GLORiA2 dataset, which can be used for the training of computer vision models aimed at detecting fish farm escapes caused by sea storms. Escaped fish can be recognized based on their appearance from pictures taken in lab conditions, when compared to farmed and wild specimens. Several images of fish specimens raised in captivity (C), farm escapes (E), and the sea (S) are provided for each of the three considered commercial species: Mediterranean sea bream (S. aurata), sea bass (D. labrax), and meagre (A. regius). The images provided are 224x224 crops of the original images, centred around the fish in the image, as preprocessed by the accompanying Python scripts. Acknowledgement: GLORiA2 is a project supported by the Biodiversity Foundation of the Ministry for the Ecological Transition, through the FEMP Pleamar Programme. It is also part of the LIFE IP INTEMARES project "Integrated, innovative and participatory management of the Natura 2000 Network in the Spanish marine environment", coordinated by the Ministry, through the Biodiversity Foundation.