no code implementations • 23 Dec 2019 • S. Balaji, T. Kavya, Natasha Sebastian
In this paper, we explore the concept of adding learn-able slope and mean shift parameters to an activation function to improve the total response region.
