Unpaired Portrait Drawing Generation via Asymmetric Cycle Mapping

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 significant imbalance of information richness between photos and drawings, existing unpaired transfer methods such as Cy-cleGAN 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). Along with localized discriminators for the eyes, nose and lips, our method well preserves all important facial features 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.

Detailed Feature Guided Generative Adversarial Pose Reconstruction Network

“…The posture with rich details make model reconstruct more realistic target face. Inspired by the study of Ran Yi et al [31], we employ [31] to obtain a sketch image of each profile face.…”

Multi-Style Artistic Portrait Drawing Generation

“…The software is the implementation for the CVPR 2020 paper in artistic portrait drawing generation [1]. The paper proposes an artistic portrait drawing generation model that learns from unpaired training data.…”

An Attention-Based Generative Adversarial Network for Producing Illustrative Sketches

“…Early approaches apply explicit computational models that produce lines and tonal expressions from an input photograph [1][2][3][4][5]. Many recent approaches apply rapidly progressing deep learning (DL) techniques [6][7][8][9][10].…”

“…These models, however, have limitations in the clear expression of salient edges, smooth tones, and vacant spaces, because they produce styles using texture transfer. Some dedicated models produce illustrative sketch styles from portraits [9,10]. Although they successfully produce visually convincing illustrative sketch images, they have limitations in the production of illustrative styles for photographs, which are not included in their training dataset.…”

“…Our second idea is to apply a relaxed cycle consistency loss (RCCL) [10] to evaluate the quality of the produced illustrative sketch styles. We reconstruct the input photograph from the produced illustrative sketch styles and compare it with the input photograph to evaluate the styles.…”