Dorta, G;
Vicente, S;
Agapito, L;
Campbell, NDF;
Prince, S;
Simpson, I;
(2017)
Laplacian pyramid of conditional variational autoencoders.
In:
(Proceedings) Proceedings of the 14th European Conference on Visual Media Production (CVMP 2017).
(pp. Article No.7).
ACM: New York, NY, USA.
Preview |
Text
a7-Dorta.pdf - Accepted Version Download (5MB) | Preview |
Abstract
Variational Autoencoders (VAE) learn a latent representation of image data that allows natural image generation and manipulation. However, they struggle to generate sharp images. To address this problem, we propose a hierarchy of VAEs analogous to a Laplacian pyramid. Each network models a single pyramid level, and is conditioned on the coarser levels. The Laplacian architecture allows for novel image editing applications that take advantage of the coarse to fine structure of the model. Our method achieves lower reconstruction error in terms of MSE, which is the loss function of the VAE and is not directly minimised in our model. Furthermore, the reconstructions generated by the proposed model are preferred over those from the VAE by human evaluators.
| Type: | Proceedings paper |
|---|---|
| Title: | Laplacian pyramid of conditional variational autoencoders |
| Event: | Proceedings of the 14th European Conference on Visual Media Production (CVMP 2017) |
| ISBN-13: | 9781450353298 |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1145/3150165.3150172 |
| Publisher version: | https://doi.org/10.1145/3150165.3150172 |
| Language: | English |
| Additional information: | This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10074876 |
Archive Staff Only
 |
View Item |
