Unpaired portrait drawing generation via asymmetric cycle mapping

Yi, Ran, Liu, Yong-Jin, Lai, Yu-Kun ORCID: https://orcid.org/0000-0002-2094-5680 and Rosin, Paul L. ORCID: https://orcid.org/0000-0002-4965-3884 2020. Unpaired portrait drawing generation via asymmetric cycle mapping. Presented at: Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, Washington, USA, 16-18 June 2020.

Preview

PDF - Accepted Post-Print Version Download (8MB) | Preview

Abstract

Portrait drawing is a common form of art with high abstraction and expressiveness. Due to its unique characteristics, existing methods achieve decent results only with paired training data, which is costly and time-consuming to obtain. In this paper, we address the problem of automatic transfer from face photos to portrait drawings with unpaired training data. We observe that due to the significantimbalanceofinformationrichnessbetweenphotosand drawings, existing unpaired transfer methods such as CycleGAN tend to embed invisible reconstruction information indiscriminately in the whole drawings, leading to important facial features partially missing in drawings. To address this problem, we propose a novel asymmetric cycle mapping that enforces the reconstruction information to be visible (by a truncation loss) and only embedded in selective facial regions (by a relaxed forward cycle-consistency loss). Alongwithlocalizeddiscriminatorsfortheeyes,nose andlips,ourmethodwellpreservesallimportantfacialfeatures in the generated portrait drawings. By introducing a style classifier and taking the style vector into account, our method can learn to generate portrait drawings in multiple styles using a single network. Extensive experiments show that our model outperforms state-of-the-art methods.

Item Type:	Conference or Workshop Item (Paper)
Status:	In Press
Schools:	Computer Science & Informatics
Funders:	The Royal Society
Date of First Compliant Deposit:	30 March 2020
Date of Acceptance:	27 February 2020
Last Modified:	26 Aug 2023 17:39
URI:	https://orca.cardiff.ac.uk/id/eprint/130656

Citation Data

Cited 31 times in Scopus. View in Scopus. Powered By Scopus® Data

Actions (repository staff only)

Edit Item