Pioneer Networks: Progressively Growing Generative Autoencoder

Ari Heljakka, Arno Solin, Juho Kannala

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussaConference article in proceedingsScientificvertaisarvioitu

14 Sitaatiot (Scopus)
235 Lataukset (Pure)

Abstrakti

We introduce a novel generative autoencoder network model that learns to encode and reconstruct images with high quality and resolution, and supports smooth random sampling from the latent space of the encoder. Generative adversarial networks (GANs) are known for their ability to simulate random high-quality images, but they cannot reconstruct existing images. Previous works have attempted to extend GANs to support such inference but, so far, have not delivered satisfactory high-quality results. Instead, we propose the Progressively Growing Generative Autoencoder (Pioneer) network which achieves high-quality reconstruction with images without requiring a GAN discriminator. We merge recent techniques for progressively building up the parts of the network with the recently introduced adversarial encoder–generator network. The ability to reconstruct input images is crucial in many real-world applications, and allows for precise intelligent manipulation of existing images. We show promising results in image synthesis and inference, with state-of-the-art results in CelebA inference tasks.

AlkuperäiskieliEnglanti
OtsikkoComputer Vision – ACCV 2018 - 14th Asian Conference on Computer Vision, Revised Selected Papers
ToimittajatGreg Mori, C.V. Jawahar, Konrad Schindler, Hongdong Li
JulkaisupaikkaSpringer, Cham
KustantajaSpringer
Sivut22-38
Sivumäärä17
ISBN (elektroninen)978-3-030-20887-5
ISBN (painettu)978-3-030-20886-8
DOI - pysyväislinkit
TilaJulkaistu - 2019
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaAsian Conference on Computer Vision - Perth, Austraalia
Kesto: 2 jouluk. 20186 jouluk. 2018
Konferenssinumero: 14

Julkaisusarja

NimiLecture notes in computer science
KustantajaSpringer Nature
Vuosikerta11361
ISSN (elektroninen)1611-3349

Conference

ConferenceAsian Conference on Computer Vision
LyhennettäACCV
Maa/AlueAustraalia
KaupunkiPerth
Ajanjakso02/12/201806/12/2018

Sormenjälki

Sukella tutkimusaiheisiin 'Pioneer Networks: Progressively Growing Generative Autoencoder'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä