Data-attention-YOLO (DAY): A comprehensive framework for mesoscale eddy identification

作者：

Highlights：

• We propose a novel framework, namely DAY, including data integration and dynamic attention modules to precisely identify mesoscale eddies.

• We devise an eddy detection module based on one-stage detecting mechanism to deeply characterize eddy features.

• The performance of our DAY framework is analyzed in comparison to other deep learning based methods with 17.3% mAP improvement.

摘要

•We propose a novel framework, namely DAY, including data integration and dynamic attention modules to precisely identify mesoscale eddies.•We devise an eddy detection module based on one-stage detecting mechanism to deeply characterize eddy features.•The performance of our DAY framework is analyzed in comparison to other deep learning based methods with 17.3% mAP improvement.

论文关键词：Mesoscale eddy identification,Attention mechanism,Data-attention-based YOLO,One-stage detection

论文评审过程：Received 22 December 2021, Revised 29 April 2022, Accepted 20 June 2022, Available online 22 June 2022, Version of Record 26 June 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.108870