Closed renmmmmmm closed 4 years ago
Does the variable s and r in Equation 1 come from different cnn backbone? Do these two network structures need to be trained separately?
Refer to the paper, it says all you need right before the equation.
Does the variable s and r in Equation 1 come from different cnn backbone? Do these two network structures need to be trained separately?