Learning GANs in simultaneous game using Sinkhorn with positive features

• Main Authors: Risman Adnan, Muchlisin Adi Saputra, Junaidillah Fadlil, Ezerman, Martianus Frederic, Muhamad Iqbal, Tjan Basaruddin
• Other Authors: School of Physical and Mathematical Sciences
• Format: Article
• Language: English
• Published: 2022
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Generative Adversarial Network; Entropy Regularized Optimal Transport
• Online Access: https://hdl.handle.net/10356/155575
• Institution: Nanyang Technological University

Entropy regularized optimal transport (EOT) distance and its symmetric normalization, known as the Sinkhorn divergence, offer smooth and continuous metrized weak-convergence distance metrics. They have excellent geometric properties and are useful to compare probability distributions in some generat...