基于深度学习的SAR图像海洋涡旋自动检测模型
刘太龙1(), 谢涛1,2,3,4(), 李建1,3,4, 王超1,5, 张雪红1,3,4
Deep learning-based automatic detection model for ocean eddies in SAR images
LIU Tailong1(), XIE Tao1,2,3,4(), LI Jian1,3,4, WANG Chao1,5, ZHANG Xuehong1,3,4

图5. 多尺度特征空间金字塔模块(MFSP)结构图

Conv代表卷积层, MaxPool2d代表二维最大池化层, EMA代表高效多尺度注意力机制, Concat代表拼接层

Fig. 5. MFSP structure diagram. Conv represents convolutional layer, MaxPool2d represents two-dimensional maximum pooling layer, EMA represents Efficient Multi-scale Attention mechanism, Concat represents concatenation layer