预测时出现错误 about ce-net HOT 6 OPEN

use strict = False mode when loading the pretrained model.
i change line 180 in file "test_cenet.py" as follows:
self.net.load_state_dict(model, strict = False)

it works !

from ce-net.

I can't find the pretrained model (CE-Net_DRIVE_auc_0.9819.th), where do you get it???

from ce-net.

from ce-net.

You should train on  the  dataset.
…
---原始邮件--- 发件人: "lzw3232"<[email protected]> 发送时间: 2020年10月22日(周四) 下午5:36 收件人: "Guzaiwang/CE-Net"<[email protected]>; 抄送: "ShenghaiLiao"<[email protected]>;"Comment"<[email protected]>; 主题: Re: [Guzaiwang/CE-Net] 预测时出现错误 (#21) I can't find the pretrained model (CE-Net_DRIVE_auc_0.9819.th), where do you get it??? — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

They release a pretrained model, which achieves 0.9819 in the AUC scor in the DRIVE dataset. I want to compare my model with CE-Net model,

from ce-net.

i don't think you should use 'strict = False', please click this link
https://blog.csdn.net/xiakejiang/article/details/93001512?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522163852170816780269813209%2522%252C%2522scm%2522%253A%252220140713.130102334..%2522%257D&request_id=163852170816780269813209&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2allbaidu_landing_v2~default-1-93001512.first_rank_v2_pc_rank_v29&utm_term=missing+keys+in+state_dict&spm=1018.2226.3001.4187

from ce-net.

i don't think you should use 'strict = False', please click this link
https://blog.csdn.net/xiakejiang/article/details/93001512?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522163852170816780269813209%2522%252C%2522scm%2522%253A%252220140713.130102334..%2522%257D&request_id=163852170816780269813209&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2allbaidu_landing_v2~default-1-93001512.first_rank_v2_pc_rank_v29&utm_term=missing+keys+in+state_dict&spm=1018.2226.3001.4187

from ce-net.