StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows

Abstract

High-quality, diverse, and photorealistic images can now be generated by unconditional GANs (e.g., StyleGAN). However, limited options exist to control the generation process using (semantic) attributes, while still preserving the quality of the output. Further, due to the entangled nature of the GAN latent space, performing edits along one attribute can easily result in unwanted changes along other attributes. In this paper, in the context of conditional exploration of entangled latent spaces, we investigate the two sub-problems of attribute-conditioned sampling and attribute-controlled editing. We present StyleFlow as a simple, effective, and robust solution to both the sub-problems by formulating conditional exploration as an instance of conditional continuous normalizing flows in the GAN latent space conditioned by attribute features. We evaluate our method using the face and the car latent space of StyleGAN, and demonstrate fine-grained disentangled edits along various attributes on both real photographs and StyleGAN generated images. For example, for faces, we vary camera pose, illumination variation, expression, facial hair, gender, and age. Finally, via extensive qualitative and quantitative comparisons, we demonstrate the superiority of StyleFlow to other concurrent works.

"Project Page https://rameenabdal.github.io/StyleFlow Video: https://youtu.be/LRAUJUn3EqQ "

References in corpus (2)

• CariGANs: Unpaired Photo-to-Caricature Translation
• MichiGAN: Multi-Input-Conditioned Hair Image Generation for Portrait Editing

Cited by in corpus (21)

• Drag Your GAN: Interactive Point-based Manipulation on the Generative Image Manifold
• SofGAN: A Portrait Image Generator with Dynamic Styling
• MatFormer: A Generative Model for Procedural Materials
• ClipFace: Text-guided Editing of Textured 3D Morphable Models
• Physically Interpretable Feature Learning and Inverse Design of Supercritical Airfoils
• On Generating Identifiable Virtual Faces
• Controllable Data Generation by Deep Learning: A Review
• Everything is There in Latent Space: Attribute Editing and Attribute Style Manipulation by StyleGAN Latent Space Exploration
• A Survey on Deep Generative 3D-aware Image Synthesis
• FreeStyleGAN: Free-view Editable Portrait Rendering with the Camera Manifold
• GANravel: User-Driven Direction Disentanglement in Generative Adversarial Networks
• Attribute-specific Control Units in StyleGAN for Fine-grained Image Manipulation
• Gaussian Splatting Decoder for 3D-aware Generative Adversarial Networks
• Stylized Face Sketch Extraction via Generative Prior with Limited Data
• FACEMUG: A Multimodal Generative and Fusion Framework for Local Facial Editing
• Text2Avatar: Text to 3D Human Avatar Generation with Codebook-Driven Body Controllable Attribute
• Design of Restricted Normalizing Flow towards Arbitrary Stochastic Policy with Computational Efficiency
• CCStereo: Audio-Visual Contextual and Contrastive Learning for Binaural Audio Generation
• ComFace: Facial Representation Learning with Synthetic Data for Comparing Faces
• A Reference-Based 3D Semantic-Aware Framework for Accurate Local Facial Attribute Editing
• SAT3D: Image-driven Semantic Attribute Transfer in 3D