Differentiable Approximation Bridges For Training Networks Containing Non-Differentiable Functions

Open in new window