thanhhungqb / pyramid-super-resolution Goto Github PK

Thanks for your contribution to the community! As shown in paperwithcode.com, the unweighted accuracy is calculated by (#corrected predictions)/(#total instances), about 80.78% for RAF-DB in your paper. But the calculation method of accuracy or weighted accuracy (about 88.98% for RAF-DB in your paper) seems not clear in the paper or in the codes. The weighted vector for the weighted accuracy is also not clear in the other related papers. I thought it was calculated by taking average over the accuracy of each expression class, but the accuracy obtained from such a calcuation is often smaller than the unweighted accuracy in the experiments. Could you please tell us how the the weighted vector for the weighted accuracy is calculated in the expriment? Looking forward to your reply. Thanks again!

in your markdown ,ferplus-vgg16-3size.json is missing , could you give me a copy？

Hello.

Thanks for your code. I've wanted to make py-script to infer your model instead of prlab.cli launches you've provided. So, I've simply wanted to load pre-trained weights to the model. If i got it right, I'm need to use model PyramidSRShare, but the model consists of several sub-networks like stn, classifier, pyramid_groups. You've provided edsr_baseline_x2-1bc95232.pt pre-trained file, which i've tried to load like in load_weights

for idx in range(len(self.multiples)):
     xn_name = 'weight_path_x{}'.format(self.multiples[idx])
     if self.config.get(xn_name, None) is not None:
         out = self.pyramid_group[idx].load_state_dict(torch.load(self.config[xn_name]), strict=True)
         print('load weights for {}'.format(xn_name), out)

As i understand, this x2 file should be loaded to pyramid_group[1] which was created
pyramid_group.append(self.make_layer_pyramid(mul, input_spec)) using mul = 2. But when i'm trying to load the weights, i'm getting "Unexpected keys" and "Missing keys" error. So, probably, i'\m doing something wrong.

Moreover, even if I'll be able to load properly weights to pyramid_group[1] there are still stn, classifier and other pyramid_group[0,2] which won't be initialized with pretrained weights. So, should there be some other pre-trained weights too?

Could you share a copy of the metadata file created for Raf-db dataset?
Or provide more details on column 5 i.e. is_test. Is this value set to False for all images?

Please I got the below error when running on colab the line : python -m prlab.cli k_fold --json_conf config/raf-db.json , I downgraded fast ai and pytorch version and still not working please advise.
Error:
File "", line 1
python -m prlab.cli k_fold --json_conf config/raf-db.json
^
SyntaxError: invalid syntax