AECRNet-MindSpore
AECRNet-MindSpore copied to clipboard
MindSpore implementation of Contrastive Learning for Compact Single Image Dehazing, CVPR2021
no moduLe name ' deconv'
请问,有没有pre-trained-model可以测试效果的?
model.py中有这样一行代码,from models.edsr_model import MeanShift,但是models文件夹中只有model.py和DCN.py,代码没有放完整吗
您好! 非常感谢提供的代码,请问这份代码是否能够复现论文里的结果? 谢谢!
对mindspore不太熟,我是使用pytorch版本的网络模型进行训练,发现模型参数大小和论文不符。您论文里面提到AECR模型的参数大小为2.61M,但是我在使用您给出的网络训练出来的模型参数为10.5M,远远大于论文中的描述。。请问能否能够解释一下原因
```python def __init__(self, input_nc, output_nc, ngf=64): ... self.block = DehazeBlock(default_conv, ngf*4, 3) ... def construct(self, x): ... x1 = self.block(x_down3) x2 = self.block(x1) x3 = self.block(x2) x4 = self.block(x3) x5...
Thanks for your excellent work! Can you release complete implementation code based Pytorch? Previous official Pytorch implementation is uncompleted , and MindSpore version is hard to train on my own...
train.py
Hello, can you provide the training code?