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
8.34k stars 1.43k forks source link

Do you have a pre-trained model of u2netp for human portrait drawing？ #322

Open zozoteacher opened 1 year ago

xuebinqin commented 1 year ago

Thanks for your interest. We didn't train the light model for human portrait segmentation. You can train it with the indicated dataset.

On Thu, Jul 21, 2022 at 8:26 PM zozoteacher @.***> wrote:

— Reply to this email directly, view it on GitHub https://github.com/xuebinqin/U-2-Net/issues/322, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORLWONCCLGUVX2NNWI3VVIIFHANCNFSM54JXNMSA . You are receiving this because you are subscribed to this thread.Message ID: @.***>

-- Xuebin Qin PhD Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage: https://xuebinqin.github.io/

zozoteacher commented 1 year ago

Do I need to train 100000 epochs?

xuebinqin commented 1 year ago

not really, it is just a maximal number for avoiding endless training. Usually you can set early stop strategy by yourself. for example, observing the convergence of training curves or the accuracy curves.

On Thu, Jul 21, 2022 at 11:53 PM zozoteacher @.***> wrote:

Do I need to train 100000 epochs?

— Reply to this email directly, view it on GitHub https://github.com/xuebinqin/U-2-Net/issues/322#issuecomment-1192246212, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORKHOKSWZBQBWTGNCGDVVJAQBANCNFSM54JXNMSA . You are receiving this because you commented.Message ID: @.***>

-- Xuebin Qin PhD Department of Computing Science University of Alberta, Edmonton, AB, Canada Homepage: https://xuebinqin.github.io/