Improving Adversarial Energy-Based Model via Diffusion Process

Cong Geng, Tian Han, Peng-Tao Jiang, Hao Zhang, Jinwei Chen, Søren Hauberg, Bo Li
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:15381-15401, 2024.

Abstract

Generative models have shown strong generation ability while efficient likelihood estimation is less explored. Energy-based models (EBMs) define a flexible energy function to parameterize unnormalized densities efficiently but are notorious for being difficult to train. Adversarial EBMs introduce a generator to form a minimax training game to avoid expensive MCMC sampling used in traditional EBMs, but a noticeable gap between adversarial EBMs and other strong generative models still exists. Inspired by diffusion-based models, we embedded EBMs into each denoising step to split a long-generated process into several smaller steps. Besides, we employ a symmetric Jeffrey divergence and introduce a variational posterior distribution for the generator’s training to address the main challenges that exist in adversarial EBMs. Our experiments show significant improvement in generation compared to existing adversarial EBMs, while also providing a useful energy function for efficient density estimation.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-geng24a, title = {Improving Adversarial Energy-Based Model via Diffusion Process}, author = {Geng, Cong and Han, Tian and Jiang, Peng-Tao and Zhang, Hao and Chen, Jinwei and Hauberg, S{\o}ren and Li, Bo}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {15381--15401}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/geng24a/geng24a.pdf}, url = {https://proceedings.mlr.press/v235/geng24a.html}, abstract = {Generative models have shown strong generation ability while efficient likelihood estimation is less explored. Energy-based models (EBMs) define a flexible energy function to parameterize unnormalized densities efficiently but are notorious for being difficult to train. Adversarial EBMs introduce a generator to form a minimax training game to avoid expensive MCMC sampling used in traditional EBMs, but a noticeable gap between adversarial EBMs and other strong generative models still exists. Inspired by diffusion-based models, we embedded EBMs into each denoising step to split a long-generated process into several smaller steps. Besides, we employ a symmetric Jeffrey divergence and introduce a variational posterior distribution for the generator’s training to address the main challenges that exist in adversarial EBMs. Our experiments show significant improvement in generation compared to existing adversarial EBMs, while also providing a useful energy function for efficient density estimation.} }
Endnote
%0 Conference Paper %T Improving Adversarial Energy-Based Model via Diffusion Process %A Cong Geng %A Tian Han %A Peng-Tao Jiang %A Hao Zhang %A Jinwei Chen %A Søren Hauberg %A Bo Li %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-geng24a %I PMLR %P 15381--15401 %U https://proceedings.mlr.press/v235/geng24a.html %V 235 %X Generative models have shown strong generation ability while efficient likelihood estimation is less explored. Energy-based models (EBMs) define a flexible energy function to parameterize unnormalized densities efficiently but are notorious for being difficult to train. Adversarial EBMs introduce a generator to form a minimax training game to avoid expensive MCMC sampling used in traditional EBMs, but a noticeable gap between adversarial EBMs and other strong generative models still exists. Inspired by diffusion-based models, we embedded EBMs into each denoising step to split a long-generated process into several smaller steps. Besides, we employ a symmetric Jeffrey divergence and introduce a variational posterior distribution for the generator’s training to address the main challenges that exist in adversarial EBMs. Our experiments show significant improvement in generation compared to existing adversarial EBMs, while also providing a useful energy function for efficient density estimation.
APA
Geng, C., Han, T., Jiang, P., Zhang, H., Chen, J., Hauberg, S. & Li, B.. (2024). Improving Adversarial Energy-Based Model via Diffusion Process. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:15381-15401 Available from https://proceedings.mlr.press/v235/geng24a.html.

Related Material