search

xuebinqin / U-2-Net

The code for our newly accepted paper in Pattern Recognition 2020: "U^2-Net: Going Deeper with Nested U-Structure for Salient Object Detection."
Apache License 2.0
8.67k stars 1.5k forks source link

why not use ssim and iou loss which used in BASNet? #228

Open Alan-Han opened 3 years ago

Alan-Han commented 3 years ago

hi qin, I open this issue to ask why not use ssim and iou loss which used in BASNet?It seems they are effective in BASBet, shown in Ablation Study part. Could you please tell me why you abandon them in U2Net, only use bce loss?

whcjb commented 3 years ago

i meet the same confusion，could you please tell the reason ？

xuebinqin commented 3 years ago

Because we found that u2net is able to capture multi-scale features well. It is not necessary to use the hybrid loss with u2net for improving the performance in terms of the existing evaluation metrics such as Fmeasure. But if your targets have rigid boundaries. We suggest to use the hybrid loss, which is able to give strong confidence around boundary regions and visually looks better.

On Mon, Oct 25, 2021 at 7:10 AM Jim Yan @.***> wrote:

i meet the same confusion，could you please �tell the reason ？

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/xuebinqin/U-2-Net/issues/228#issuecomment-950486611, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORMPDIEEQYY7L6C3VLDUITDDFANCNFSM5AAY4L3Q . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

-- Xuebin Qin PhD Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage:https://webdocs.cs.ualberta.ca/~xuebin/