This is a SE_DenseNet which contains a senet (Squeeze-and-Excitation Networks by Jie Hu, Li Shen, and Gang Sun) module, written in Pytorch, train, and eval codes have been released.
Hi) Thank you for your great work!
Could you show an amount of model parameters? I'm trying to implement SE-DenseNet from the scratch for my diploma work and I am not sure that my SE implementation is correct (it adds more than 3 millions parameters...)
Hi, thanks for sharing your experiment results. I checked and found that you may have some redundant code in the _Dense layer that adds thr seblock in of convolution. You added it in loop (for) and after first convolution. Why do you add seblock in _Dense_layer again? Thanks
hi
I see in se_densenet_full.py, it use model.load_state_dict(state_dict, strict=is_strict)
And in test_se_densenet.py, it use
model = se_densenet121(pretrained=pretrained)
net_state_dict = {key: value for key, value in model_zoo.load_url("https://download.pytorch.org/models/densenet121-a639ec97.pth").items()}
model.load_state_dict(net_state_dict, strict=False)
why?