eduzhai > Applied Sciences > Engineering >

Mixup-CAM Weakly-supervised Semantic Segmentation via Uncertainty Regularization

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 13 pages

Abstract: Obtaining object response maps is one important step to achieveweakly-supervised semantic segmentation using image-level labels. However,existing methods rely on the classification task, which could result in aresponse map only attending on discriminative object regions as the networkdoes not need to see the entire object for optimizing the classification loss.To tackle this issue, we propose a principled and end-to-end train-ableframework to allow the network to pay attention to other parts of the object,while producing a more complete and uniform response map. Specifically, weintroduce the mixup data augmentation scheme into the classification networkand design two uncertainty regularization terms to better interact with themixup strategy. In experiments, we conduct extensive analysis to demonstratethe proposed method and show favorable performance against state-of-the-artapproaches.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×