Firstly, channel attention map(CAM) and spatial attention map(SAM) were embedded in the CycleGAN model to enhance the feature extraction ability of the model. Secondly, the Weight Demodulation(WD) mechanism was introduced to fix feature artifacts and white spots, further improving the quality of ...
为解决上述问题,本文以CycleGAN为基础,将钢材表面缺陷图像生成看作图像风格迁移任务的一种变形,[JP+1]运用CycleGAN网络,实现不同种类钢材表面缺陷特征之间的风格迁移,然后通过添加注意力机制、引入权重解调机制和形状一致性损失来改进模型,提出一种基于改进CycleGAN的钢材表面缺陷图像生成模型。 1相关工作 1.1生成对抗网络(...
针对工业钢材表面缺陷检测过程中存在的样本采集困难,成本较高,以及缺陷种类较多难以覆盖全部导致的小样本问题,提出一种改进循环生成对抗网络(cycle-consistentgenerativeadversarialnetworks,CycleGAN)的钢材表面缺陷图像生成方法.首先,将通道注意力(classactivationmap,CAM)和空间注意力(spatialattentionmap,SAM)机制嵌入到CycleGAN...