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
8.43k stars 1.45k forks source link

Should I flip the training dataset images or does your code do that? #331

Open alcinopitchprint opened 1 year ago

alcinopitchprint commented 1 year ago

Hi,

Thank you for your work.

I would just like to know if I should flip my training dataset before training or does your training process flip the images.

Just not too clear on this part of the paper: "We augment this dataset by horizontal flipping to obtain 21106 training images offline"

Thank you :-)

xuebinqin commented 1 year ago

In this paper, the images are flipped offline, which means before the training process, but you can also flip them online.

On Wed, Sep 7, 2022 at 1:29 AM Alcino Van Rooyen @.***> wrote:

Hi,

Thank you for your work.

I would just like to know if I should flip my training dataset before training or does your training process flip the images.

Just not too clear on this part of the paper: "We augment this dataset by horizontal flipping to obtain 21106 training images offline"

Thank you :-)

— Reply to this email directly, view it on GitHub https://github.com/xuebinqin/U-2-Net/issues/331, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORKRZU7IBZRN7PSYPLTV5BG47ANCNFSM6AAAAAAQGQ4HIE . 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/

alcinopitchprint commented 1 year ago

So what do you mean by online?