Li, Kun, Liu, Jingying, Lai, Yukun ORCID: https://orcid.org/0000-0002-2094-5680 and Yang, Jingyu 2019. Generating 3D faces using multi-column graph convolutional networks. Computer Graphics Forum 38 (7) , pp. 215-224. |
Preview |
PDF
- Accepted Post-Print Version
Download (5MB) | Preview |
Abstract
In this work, we introduce multi-column graph convolutional networks (MGCNs), a deep generative model for 3D mesh surfaces that effectively learns a non-linear facial representation. We perform spectral decomposition of meshes and apply convolutions directly in the frequency domain. Our network architecture involves multiple columns of graph convolutional networks (GCNs), namely large GCN (L-GCN), medium GCN (M-GCN) and small GCN (S-GCN), with different filter sizes to extract features at different scales. L-GCN is more useful to extract large-scale features, whereas S-GCN is effective for extracting subtle and fine-grained features, and M-GCN captures information in between. Therefore, to obtain a high-quality representation, we propose a selective fusion method that adaptively integrates these three kinds of information. Spatially non-local relationships are also exploited through a self-attention mechanism to further improve the representation ability in the latent vector space. Through extensive experiments, we demonstrate the superiority of our end-to-end framework in improving the accuracy of 3D face reconstruction. Moreover, with the help of variational inference, our model has excellent generating ability.
Item Type: | Article |
---|---|
Date Type: | Publication |
Status: | Published |
Schools: | Computer Science & Informatics |
Additional Information: | A special issue of Pacific Graphics 2019 paper that will be published in Computer Graphics Forum. |
Publisher: | Wiley |
ISSN: | 0167-7055 |
Date of First Compliant Deposit: | 1 October 2019 |
Date of Acceptance: | 26 August 2019 |
Last Modified: | 25 Nov 2024 14:30 |
URI: | https://orca.cardiff.ac.uk/id/eprint/125794 |
Citation Data
Cited 6 times in Scopus. View in Scopus. Powered By Scopus® Data
Actions (repository staff only)
Edit Item |