Deep Automodulators

Ari Heljakka, Yuxin Hou, Juho Kannala, Arno Solin

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review


We introduce a new category of generative autoencoders called automodulators. These networks can faithfully reproduce individual real-world input images like regular autoencoders, but also generate a fused sample from an arbitrary combination of several such images, allowing instantaneous "style-mixing" and other new applications. An automodulator decouples the data flow of decoder operations from statistical properties thereof and uses the latent vector to modulate the former by the latter, with a principled approach for mutual disentanglement of decoder layers. Prior work has explored similar decoder architecture with GANs, but their focus has been on random sampling. A corresponding autoencoder could operate on real input images. For the first time, we show how to train such a general-purpose model with sharp outputs in high resolution, using novel training techniques, demonstrated on four image data sets. Besides style-mixing, we show state-of-the-art results in autoencoder comparison, and visual image quality nearly indistinguishable from state-of-the-art GANs. We expect the automodulator variants to become a useful building block for image applications and other data domains.
Original languageEnglish
Title of host publicationThirty-fourth Conference on Neural Information Processing Systems
PublisherMorgan Kaufmann Publishers
Publication statusPublished - 2020
MoE publication typeA4 Conference publication
EventConference on Neural Information Processing Systems - Virtual, Vancouver, Canada
Duration: 6 Dec 202012 Dec 2020
Conference number: 34

Publication series

NameAdvances in Neural Information Processing Systems
PublisherMorgan Kaufmann Publishers
ISSN (Electronic)1049-5258


ConferenceConference on Neural Information Processing Systems
Abbreviated titleNeurIPS


Dive into the research topics of 'Deep Automodulators'. Together they form a unique fingerprint.

Cite this