dc.citation.conferencePlace |
GE |
- |
dc.citation.conferencePlace |
Munich |
- |
dc.citation.endPage |
678 |
- |
dc.citation.startPage |
662 |
- |
dc.citation.title |
European Conference on Computer Vision |
- |
dc.contributor.author |
Gokaslan, Aaron |
- |
dc.contributor.author |
Ramanujan, Vivek |
- |
dc.contributor.author |
Ritchie, Daniel |
- |
dc.contributor.author |
Kim, Kwang In |
- |
dc.contributor.author |
Tompkin, James |
- |
dc.date.accessioned |
2024-02-01T01:36:28Z |
- |
dc.date.available |
2024-02-01T01:36:28Z |
- |
dc.date.created |
2019-02-28 |
- |
dc.date.issued |
2018-09-08 |
- |
dc.description.abstract |
Unsupervised image-to-image translation techniques are able to map local texture between two domains, but they are typically unsuccessful when the domains require larger shape change. Inspired by semantic segmentation, we introduce a discriminator with dilated convolutions that is able to use information from across the entire image to train a more context-aware generator. This is coupled with a multi-scale perceptual loss that is better able to represent error in the underlying shape of objects. We demonstrate that this design is more capable of representing shape deformation in a challenging toy dataset, plus in complex mappings with significant dataset variation between humans, dolls, and anime faces, and between cats and dogs. © Springer Nature Switzerland AG 2018. |
- |
dc.identifier.bibliographicCitation |
European Conference on Computer Vision, pp.662 - 678 |
- |
dc.identifier.doi |
10.1007/978-3-030-01258-8_40 |
- |
dc.identifier.issn |
0302-9743 |
- |
dc.identifier.scopusid |
2-s2.0-85055128063 |
- |
dc.identifier.uri |
https://scholarworks.unist.ac.kr/handle/201301/80948 |
- |
dc.identifier.url |
https://link.springer.com/chapter/10.1007%2F978-3-030-01258-8_40 |
- |
dc.language |
영어 |
- |
dc.publisher |
ECCV 2018 |
- |
dc.title |
Improving shape deformation in unsupervised image-to-image translation |
- |
dc.type |
Conference Paper |
- |
dc.date.conferenceDate |
2018-09-08 |
- |