Non-parametric priors for GANs accepted at ICML 2019

Moving away from standard parametric priors, our approach using non-parametric priors helps GANs counter distribution-mismatch problems. Just been accepted at the Intl. Conference on Machine Learning (ICML) 2019. Congrats to Rajhans Singh and team. Opens potentially new ways to enforce constraints on GAN outputs by designing priors more deliberately. Paper and code coming soon.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s