Online-Ressource | |
Verfasst von: | Ardizzone, Lynton [VerfasserIn] |
Kruse, Jakob [VerfasserIn] | |
Lüth, Carsten [VerfasserIn] | |
Bracher, Niels [VerfasserIn] | |
Rother, Carsten [VerfasserIn] | |
Köthe, Ullrich [VerfasserIn] | |
Titel: | Conditional invertible neural networks for diverse image-to-image translation |
Verf.angabe: | Lynton Ardizzone, Jakob Kruse, Carsten Lüth, Niels Bracher, Carsten Rother, Ullrich Köthe |
E-Jahr: | 2021 |
Jahr: | 5 May 2021 |
Umfang: | 15 S. |
Fussnoten: | Gesehen am 28.09.2022 |
Titel Quelle: | Enthalten in: De.arxiv.org |
Ort Quelle: | [S.l.] : Arxiv.org, 1991 |
Jahr Quelle: | 2021 |
Band/Heft Quelle: | (2021), Artikel-ID 2105.02104, Seite 1-15 |
Abstract: | We introduce a new architecture called a conditional invertible neural network (cINN), and use it to address the task of diverse image-to-image translation for natural images. This is not easily possible with existing INN models due to some fundamental limitations. The cINN combines the purely generative INN model with an unconstrained feed-forward network, which efficiently preprocesses the conditioning image into maximally informative features. All parameters of a cINN are jointly optimized with a stable, maximum likelihood-based training procedure. Even though INN-based models have received far less attention in the literature than GANs, they have been shown to have some remarkable properties absent in GANs, e.g. apparent immunity to mode collapse. We find that our cINNs leverage these properties for image-to-image translation, demonstrated on day to night translation and image colorization. Furthermore, we take advantage of our bidirectional cINN architecture to explore and manipulate emergent properties of the latent space, such as changing the image style in an intuitive way. |
DOI: | doi:10.48550/arXiv.2105.02104 |
URL: | Bitte beachten Sie: Dies ist ein Bibliographieeintrag. Ein Volltextzugriff für Mitglieder der Universität besteht hier nur, falls für die entsprechende Zeitschrift/den entsprechenden Sammelband ein Abonnement besteht oder es sich um einen OpenAccess-Titel handelt. Volltext: https://doi.org/10.48550/arXiv.2105.02104 |
Volltext: http://arxiv.org/abs/2105.02104 | |
DOI: https://doi.org/10.48550/arXiv.2105.02104 | |
Datenträger: | Online-Ressource |
Sprache: | eng |
Sach-SW: | 68T01 |
Computer Science - Artificial Intelligence | |
Computer Science - Computer Vision and Pattern Recognition | |
K10plus-PPN: | 1817338129 |
Verknüpfungen: | → Sammelwerk |