search

xuebinqin / BASNet

Code for CVPR 2019 paper. BASNet: Boundary-Aware Salient Object Detection
MIT License
1.37k stars 250 forks source link

I wanna ask how long time you take to train the model on DUTS, how many epoches are ok, and the rough loss value after convergence. #37

Closed Starboy-at-earth closed 4 years ago

xuebinqin commented 4 years ago

please refer to the section 4.2 of our paper:https://webdocs.cs.ualberta.ca/~xuebin/BASNet.pdf

Starboy-at-earth commented 4 years ago

Thank you for your assistance. I have another question. Why do not you load the loss fuction to cuda for acceleration？Can pytorch acclerate it？

---Original--- From: "Xuebin Qin"<notifications@github.com> Date: Thu, Feb 20, 2020 14:15 PM To: "NathanUA/BASNet"<BASNet@noreply.github.com>; Cc: "Starboy-at-earth"<2667004002@qq.com>;"Author"<author@noreply.github.com>; Subject: Re: [NathanUA/BASNet] I wanna ask how long time you take to train the model on DUTS, how many epoches are ok, and the rough loss value after convergence. (#37)

please refer to the section 4.2 of our paper:https://webdocs.cs.ualberta.ca/~xuebin/BASNet.pdf

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

xuebinqin commented 4 years ago

I think the loss has already been accelerated and ran on GPU.

On Sat, Feb 22, 2020 at 3:59 AM Starboy-at-earth notifications@github.com wrote:

Thank you for your assistance. I have another question. Why do not you load the loss fuction to cuda for acceleration？Can pytorch acclerate it？

---Original--- From: "Xuebin Qin"<notifications@github.com> Date: Thu, Feb 20, 2020 14:15 PM To: "NathanUA/BASNet"<BASNet@noreply.github.com>; Cc: "Starboy-at-earth"<2667004002@qq.com>;"Author"< author@noreply.github.com>; Subject: Re: [NathanUA/BASNet] I wanna ask how long time you take to train the model on DUTS, how many epoches are ok, and the rough loss value after convergence. (#37)

please refer to the section 4.2 of our paper: https://webdocs.cs.ualberta.ca/~xuebin/BASNet.pdf

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/NathanUA/BASNet/issues/37?email_source=notifications&email_token=ADSGORMHZVNUXNJN2EGRZX3REEAPNA5CNFSM4KWNH2QKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEMU5L3A#issuecomment-589944300, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADSGORI2W6AAMM2V5NUJBF3REEAPNANCNFSM4KWNH2QA .

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

Starboy-at-earth commented 4 years ago

Thank you for your patience. The learnt experience in tf misleads me (embarassingly new to pytorch). I have found it. 

---Original--- From: "Xuebin Qin"<notifications@github.com> Date: Sun, Feb 23, 2020 02:08 AM To: "NathanUA/BASNet"<BASNet@noreply.github.com>; Cc: "Starboy-at-earth"<2667004002@qq.com>;"Author"<author@noreply.github.com>; Subject: Re: [NathanUA/BASNet] I wanna ask how long time you take to train the model on DUTS, how many epoches are ok, and the rough loss value after convergence. (#37)

I think the loss has already been accelerated and ran on GPU.

On Sat, Feb 22, 2020 at 3:59 AM Starboy-at-earth <notifications@github.com> wrote:

> Thank you for your assistance. I have another question. Why do not you > load the loss fuction to cuda for acceleration？Can pytorch acclerate it？ > > > > ---Original--- > From: "Xuebin Qin"<notifications@github.com&gt; > Date: Thu, Feb 20, 2020 14:15 PM > To: "NathanUA/BASNet"<BASNet@noreply.github.com&gt;; > Cc: "Starboy-at-earth"<2667004002@qq.com&gt;;"Author"< > author@noreply.github.com&gt;; > Subject: Re: [NathanUA/BASNet] I wanna ask how long time you take to train > the model on DUTS, how many epoches are ok, and the rough loss value after > convergence. (#37) > > > > please refer to the section 4.2 of our paper: > https://webdocs.cs.ualberta.ca/~xuebin/BASNet.pdf > > — > You are receiving this because you authored the thread. > Reply to this email directly, view it on GitHub, or unsubscribe. > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub > <https://github.com/NathanUA/BASNet/issues/37?email_source=notifications&amp;email_token=ADSGORMHZVNUXNJN2EGRZX3REEAPNA5CNFSM4KWNH2QKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEMU5L3A#issuecomment-589944300&gt;, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ADSGORI2W6AAMM2V5NUJBF3REEAPNANCNFSM4KWNH2QA&gt; > . >

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

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.