mczhuge / icon Goto Github PK

😃

hi, I have read your code, and I find a mistake in ICE modules L165.

in2 = F.interpolate(in1, size=in1.size()[2:], mode='bilinear')

I think it should correct like below:

in2 = F.interpolate(in2, size=in1.size()[2:], mode='bilinear')

Hi, I trained with the following config and I did not change the code
python main/train.py --model 'ICON-S' --dataset '/data2/lanjun/dataset/sod_dataset/datasets/DUTS/Train' --lr 2e-3 --decay 2e-4 --momen 0.9 --batchsize 6 --loss 'CPR' --savepath 'checkpoint/finetune/ICON-S/' --valid True

but I get the following result，it doesn't match the result in the paper. Did I do something wrong? need help
Method:ICON-valid-59,Dataset:ECSSD||Smeasure:0.927; meanEm:0.954; wFmeasure:0.911; MAE:0.032; fnr:0.065||adpEm:0.955; meanEm:0.954; maxEm:0.963; adpFm:0.918; meanFm:0.916; maxFm:0.941
Method:ICON-valid-59,Dataset:PASCAL-S||Smeasure:0.875; meanEm:0.912; wFmeasure:0.834; MAE:0.054; fnr:0.115||adpEm:0.908; meanEm:0.912; maxEm:0.921; adpFm:0.843; meanFm:0.849; maxFm:0.872
Method:ICON-valid-60,Dataset:ECSSD||Smeasure:0.926; meanEm:0.953; wFmeasure:0.911; MAE:0.032; fnr:0.069||adpEm:0.955; meanEm:0.953; maxEm:0.962; adpFm:0.919; meanFm:0.917; maxFm:0.94
Method:ICON-valid-60,Dataset:PASCAL-S||Smeasure:0.874; meanEm:0.91; wFmeasure:0.833; MAE:0.055; fnr:0.122||adpEm:0.907; meanEm:0.91; maxEm:0.919; adpFm:0.843; meanFm:0.849; maxFm:0.87

Hi, I have tested with the ResNet50 weights and dataset provided by GitHub and have not modified the code.
However, I got the following results, which are inconsistent with the paper's results.

Sm:0.812 maxFm:0.743 MAE:0.062 adpFm:0.735 meanFm:0.726 || Data:[_ICON_git/DUT-OMRON]]
Sm:0.854 maxFm:0.824 MAE:0.047 adpFm:0.794 meanFm:0.804 || Data:[_ICON_git/DUTS]]
Sm:0.886 maxFm:0.889 MAE:0.045 adpFm:0.872 meanFm:0.865 || Data:[_ICON_git/HKU-IS]]
Sm:0.845 maxFm:0.833 MAE:0.072 adpFm:0.803 meanFm:0.814 || Data:[_ICON_git/PASCAL-S]]
Sm:0.903 maxFm:0.911 MAE:0.047 adpFm:0.889 meanFm:0.890 || Data:[_ICON_git/ECSSD]]

hello, great work! I wonder how you find the data for plotting PR curves of other methods. I would like to plot the PR curves of our method and other methods for comparison, but I could not find the data for plotting the curves of other methods.

Thanks for your great work. I downloaded the predicted saliency map of ICON-R. However, there are only 4445 predicted saliency maps in HKU-IS, instead of 4447. Would you like to provide the full predicted saliency maps of HKU-IS. Thank you very much.

Traceback (most recent call last):
File "main/train.py", line 226, in
train(dataset, parser)
File "main/train.py", line 132, in train
out2, out3, out4, out5, pose = net(image)
File "/usr/local/anaconda3/envs/u2/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "/home/senlinren/LR/ICON-main/main/model/icon/icon.py", line 221, in forward
x1 = self.dfa1(x1)
File "/usr/local/anaconda3/envs/u2/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "/home/senlinren/LR/ICON-main/main/model/icon/modules.py", line 134, in forward
p3 = self.atrConv(f)
File "/usr/local/anaconda3/envs/u2/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "/usr/local/anaconda3/envs/u2/lib/python3.7/site-packages/torch/nn/modules/container.py", line 204, in forward
input = module(input)
File "/usr/local/anaconda3/envs/u2/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "/usr/local/anaconda3/envs/u2/lib/python3.7/site-packages/torch/nn/modules/activation.py", line 1240, in forward
return F.prelu(input, self.weight)
RuntimeError: expected scalar type Half but found Float

I retrained your model under the default settings, and got the following results, which are inconsistent with the paper's results:

Method:ICON-S,Dataset:ECSSD||Smeasure:0.92; meanEm:0.949; wFmeasure:0.898; MAE:0.036; adpEm:0.947; maxEm:0.958; adpFm:0.907; meanFm:0.909; maxFm:0.934; fnr:0.053
Method:ICON-S,Dataset:PASCALS||Smeasure:0.833; meanEm:0.866; wFmeasure:0.773; MAE:0.085; adpEm:0.863; maxEm:0.882; adpFm:0.792; meanFm:0.798; maxFm:0.829; fnr:0.098
Method:ICON-S,Dataset:DUTS-TE||Smeasure:0.85; meanEm:0.882; wFmeasure:0.76; MAE:0.057; adpEm:0.867; maxEm:0.906; adpFm:0.761; meanFm:0.786; maxFm:0.834; fnr:0.08
Method:ICON-S,Dataset:HKU-IS||Smeasure:0.912; meanEm:0.947; wFmeasure:0.881; MAE:0.032; adpEm:0.948; maxEm:0.958; adpFm:0.887; meanFm:0.893; maxFm:0.923; fnr:0.063
Method:ICON-S,Dataset:DUT-OMRON||Smeasure:0.818; meanEm:0.844; wFmeasure:0.712; MAE:0.078; adpEm:0.838; maxEm:0.876; adpFm:0.72; meanFm:0.739; maxFm:0.786; fnr:0.093
Method:ICON-S,Dataset:SOD||Smeasure:0.806; meanEm:0.851; wFmeasure:0.768; MAE:0.098; adpEm:0.847; maxEm:0.864; adpFm:0.812; meanFm:0.819; maxFm:0.829; fnr:0.24

Can you help me to solve it? Thank you very much!

dear author:
训练集和验证集如何划分呢?或者说训练集和验证集的大小比是多少

Hi,
Is it possible to upload the predicted saliency map into Google Drive?
I can't download them.

File "D:/pydemo/ICON/main/train.py", line 226, in
train(dataset, parser)
File "D:/pydemo/ICON/main/train.py", line 130, in train
for step, (image, mask) in enumerate(loader):

TypeError: 'NoneType' object is not callable

Hi,
could be possible to host the checkpoints on a different place?
Alessio