Proximal Mapping for Deep Regularization
NIPS 2020, 2020.
We proposed using proximal mapping as a new primitive in deep networks to explicitly encode the prior for end-to-end training
Underpinning the success of deep learning is effective regularizations that allow a variety of priors in data to be modeled. For example, robustness to adversarial perturbations, and correlations between multiple modalities. However, most regularizers are specified in terms of hidden layer outputs, which are not themselves optimization ...More
PPT (Upload PPT)