PS-StyleGAN: Illustrative Portrait Sketching using Attention-Based Style Adaptation
Preprint
Abstract
Portrait sketching involves capturing identity specific attributes of a real face with abstract lines and shades. Unlike photo-realistic images, a good portrait sketch generation method needs selective attention to detail, making the problem challenging. This paper introduces \textbf{Portrait Sketching StyleGAN (PS-StyleGAN)}, a style transfer approach tailored for portrait sketch synthesis. We leverage the semantic $W+$ latent space of StyleGAN to generate portrait sketches, allowing us to make meaningful edits, like pose and expression alterations, without compromising identity. To achieve this, we propose the use of Attentive Affine transform blocks in our architecture, and a training strategy that allows us to change StyleGAN's output without finetuning it. These blocks learn to modify style latent code by paying attention to both content and style latent features, allowing us to adapt the outputs of StyleGAN in an inversion-consistent manner. Our approach uses only a few paired examples ($\sim 100$) to model a style and has a short training time. We demonstrate PS-StyleGAN's superiority over the current state-of-the-art methods on various datasets, qualitatively and quantitatively.
Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- CLIP4Sketch: Enhancing Sketch to Mugshot Matching through Dataset Augmentation using Diffusion Models2024
- JoJoGAN: One Shot Face Stylization2022
- Facial-sketch Synthesis: A New ChallengeMachine Intelligence Research, 2022
- Why Do Line Drawings Work? A Realism HypothesisPerception, 2020
- Style and abstraction in portrait sketchingACM Transactions on Graphics, 2013
- Face Photo-Sketch Synthesis and RecognitionIEEE Transactions on Pattern Analysis and Machine Intelligence, 2008
- Mechanisms of Face PerceptionAnnual Review of Neuroscience, 2008
- Ridge-valley lines on meshes via implicit surface fittingACM Transactions on Graphics, 2004
- Suggestive contours for conveying shapeACM Transactions on Graphics, 2003
- Surface versus edge-based determinants of visual recognitionCognitive Psychology, 1988