WithWheat
WithWheat
Actually, this two implementations have different computational complexity. BTW, I wonder that the computational complexity and accuracy reported in the paper is based on which implementation? Thanks.
Okay, I see. You mean this implementation splits the first 1x1 convolution in the original bottleneck into two 1x1 convolutions. In this way, two BN+ReLU layers are required. Thanks for...
@fac33 @qizhou000 @Hunter-Murphy @flag16 Hi, guys! Just change the 'dir_data' in 'option.py' correctly then this bug will be fixed. For example, you can do it like this, '/home/aaa/bbb/DIV2K/'--->'/home/aaa/bbb/' Hope this...