Emerging Convolutions for Generative Normalizing Flows

Emiel Hoogeboom, Rianne Van Den Berg, Max Welling
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:2771-2780, 2019.

Abstract

Generative flows are attractive because they admit exact likelihood optimization and efficient image synthesis. Recently, Kingma & Dhariwal (2018) demonstrated with Glow that generative flows are capable of generating high quality images. We generalize the 1 {\texttimes} 1 convolutions proposed in Glow to invertible d {\texttimes} d convolutions, which are more flexible since they operate on both channel and spatial axes. We propose two methods to produce invertible convolutions, that have receptive fields identical to standard convolutions: Emerging convolutions are obtained by chaining specific autoregressive convolutions, and periodic convolutions are decoupled in the frequency domain. Our experiments show that the flexibility of d {\texttimes} d convolutions significantly improves the performance of generative flow models on galaxy images, CIFAR10 and ImageNet.

Cite this Paper


BibTeX
@InProceedings{pmlr-v97-hoogeboom19a, title = {Emerging Convolutions for Generative Normalizing Flows}, author = {Hoogeboom, Emiel and Van Den Berg, Rianne and Welling, Max}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {2771--2780}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/hoogeboom19a/hoogeboom19a.pdf}, url = {https://proceedings.mlr.press/v97/hoogeboom19a.html}, abstract = {Generative flows are attractive because they admit exact likelihood optimization and efficient image synthesis. Recently, Kingma & Dhariwal (2018) demonstrated with Glow that generative flows are capable of generating high quality images. We generalize the 1 {\texttimes} 1 convolutions proposed in Glow to invertible d {\texttimes} d convolutions, which are more flexible since they operate on both channel and spatial axes. We propose two methods to produce invertible convolutions, that have receptive fields identical to standard convolutions: Emerging convolutions are obtained by chaining specific autoregressive convolutions, and periodic convolutions are decoupled in the frequency domain. Our experiments show that the flexibility of d {\texttimes} d convolutions significantly improves the performance of generative flow models on galaxy images, CIFAR10 and ImageNet.} }
Endnote
%0 Conference Paper %T Emerging Convolutions for Generative Normalizing Flows %A Emiel Hoogeboom %A Rianne Van Den Berg %A Max Welling %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-hoogeboom19a %I PMLR %P 2771--2780 %U https://proceedings.mlr.press/v97/hoogeboom19a.html %V 97 %X Generative flows are attractive because they admit exact likelihood optimization and efficient image synthesis. Recently, Kingma & Dhariwal (2018) demonstrated with Glow that generative flows are capable of generating high quality images. We generalize the 1 {\texttimes} 1 convolutions proposed in Glow to invertible d {\texttimes} d convolutions, which are more flexible since they operate on both channel and spatial axes. We propose two methods to produce invertible convolutions, that have receptive fields identical to standard convolutions: Emerging convolutions are obtained by chaining specific autoregressive convolutions, and periodic convolutions are decoupled in the frequency domain. Our experiments show that the flexibility of d {\texttimes} d convolutions significantly improves the performance of generative flow models on galaxy images, CIFAR10 and ImageNet.
APA
Hoogeboom, E., Van Den Berg, R. & Welling, M.. (2019). Emerging Convolutions for Generative Normalizing Flows. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:2771-2780 Available from https://proceedings.mlr.press/v97/hoogeboom19a.html.

Related Material