two classification tasks with output channel = 1
bobopit opened this issue · comments
Hello, thank you for your excellent work. I have a question about that when the output channels of my model is 1 for 2 classification tasks, is rmi loss meaningful?
It is a special case for multi-class classification, RMI may also work for it (output should also have a similar structure to ground truth).
It is a special case for multi-class classification, RMI may also work for it (output should also have a similar structure to ground truth).
You said in your paper: "We adopt the sigmoid operation rather than softmax operation to get predicted probabilities. This is
because RMI is calculated channel-wise, we do not want to introduce interference between channels.", But in above case, the channel in output is 1, how to do calculate "channel-wise".
Sigmoid is element-wise. You can also use it here.
Sigmoid is element-wise. You can also use it here.
thanks for your reply!