Learning Discriminators as Energy Networks in Adversarial Learning

Pingbo Pan, Yan Yan, Tianbao Yang, Yi Yang

Abstract: We propose a novel adversarial learning framework in this work. Existing adversarial learning methods involve two separate networks, i.e., the structured prediction models and the discriminative models, in the training. The information captured by discriminative models complements that in the structured prediction models, but few existing researches have studied on utilizing such information to improve structured prediction models at the inference stage. In this work, we propose to refine the predictions of structured prediction models by effectively integrating discriminative models into the prediction. Discriminative models are treated as energy-based models. Similar to the adversarial learning, discriminative models are trained to estimate scores which measure the quality of predicted outputs, while structured prediction models are trained to predict contrastive outputs with maximal energy scores. In this way, the gradient vanishing problem is ameliorated, and thus we are able to perform inference by following the ascent gradient directions of discriminative models to refine structured prediction models. The proposed method is able to handle a range of tasks, \emph{e.g.}, multi-label classification and image segmentation. Empirical results on these two tasks validate the effectiveness of our learning method.

Keywords: adversarial learning, structured prediction, energy networks

TL;DR: We propose a novel adversarial learning framework for structured prediction, in which discriminative models can be used to refine structured prediction models at the inference stage.

Data: [LFW](https://paperswithcode.com/dataset/lfw)