S. Lopez-Marcano

Improved accuracy for automated counting of a fish in baited underwater videos for stock assessment

This paper improves automated counting of fish in BRUVS videos by optimizing confidence thresholds, Seq-NMS, and correction equations to achieve high accuracy across fish abundances.

Deep learning for automated analysis of fish abundance: the benefits of training across multiple habitats

This study trains deep learning models to detect luderick fish in coastal habitats, finding that models trained on multiple habitats perform better across different environments.

Automating the analysis of fish abundance using object detection: optimising animal ecology with deep learning

This study demonstrates that deep learning outperforms human experts and citizen scientists in accurately and efficiently quantifying fish abundance from underwater footage.