Oceanic mesoscale eddy detection and convolutional neural network complexity

Oliverio J. Santana, Daniel Hernández-Sosa, Ryan N. Smith

Abstract: Highlights•Increasingly complex CNN architectures are being proposed for eddy detection.•We prove that excessively complex CNN designs are not needed to solve this problem.•Our proposal improves state-of-the-art designs and requires to train much less parameters.•Our data augmentation method improves performance without artificial transformations.•Our model is less sensitive to timely variations and detects eddies that traditional models would miss.