English

Maximum Entropy Generators for Energy-Based Models

Machine Learning 2019-05-29 v2 Artificial Intelligence Machine Learning

Abstract

Maximum likelihood estimation of energy-based models is a challenging problem due to the intractability of the log-likelihood gradient. In this work, we propose learning both the energy function and an amortized approximate sampling mechanism using a neural generator network, which provides an efficient approximation of the log-likelihood gradient. The resulting objective requires maximizing entropy of the generated samples, which we perform using recently proposed nonparametric mutual information estimators. Finally, to stabilize the resulting adversarial game, we use a zero-centered gradient penalty derived as a necessary condition from the score matching literature. The proposed technique can generate sharp images with Inception and FID scores competitive with recent GAN techniques, does not suffer from mode collapse, and is competitive with state-of-the-art anomaly detection techniques.

Keywords

Cite

@article{arxiv.1901.08508,
  title  = {Maximum Entropy Generators for Energy-Based Models},
  author = {Rithesh Kumar and Sherjil Ozair and Anirudh Goyal and Aaron Courville and Yoshua Bengio},
  journal= {arXiv preprint arXiv:1901.08508},
  year   = {2019}
}