On Characterizing GAN Convergence Through Proximal Duality Gap
Sahil Sidheekh 1 Aroof Aimen 1 Narayanan C. Krishnan 1
Abstract most popular and successful of innovations for learning data
Despite the accomplishments of Generative Ad- distributions. A GAN formulates the generative modeling
versarial Networks (GANs) in modeling data dis- problem as a zero-sum game between two agents - a Dis-
tributions, training them remains a challen ...


雷达卡




京公网安备 11010802022788号







