CBAM-Unet++:easier to find the target with the attention module "CBAM"
Zhengxuan Zhao, Kaixu Chen, Satoshi Yamane
Abstract
There are already many methods based on U-net, however, due to the paricularity of medical images, we need to pay more attention to the target area to perform more detailed segmentation. In this paper, we present a CBAM-Unet++ module, which a more targeted architecture for medical image segmentation. It combines Unet++ and Convolutional block attention module to make it easier for architecture to ignore irrelevant background information, thereby paying more attention to the parts that we want to have.
Topics & Concepts
Computer scienceBlock (permutation group theory)SegmentationArchitectureComputer visionImage segmentationArtificial intelligenceMathematicsGeometryVisual artsArtAdvanced Neural Network ApplicationsMedical Image Segmentation TechniquesAdvanced Image and Video Retrieval Techniques