StyleHumanCLIP: Text-guided Garment Manipulation for StyleGAN-Human
This addresses the problem of precise text-guided editing for garments in human images, which is incremental over existing StyleGAN-based methods.
The paper tackles text-guided garment manipulation in full-body human images generated by StyleGAN, proposing an attention-based latent code mapper that enables more disentangled control than existing methods. The result shows the method can control generated images more faithfully to given texts, as confirmed by quantitative and qualitative evaluations.
This paper tackles text-guided control of StyleGAN for editing garments in full-body human images. Existing StyleGAN-based methods suffer from handling the rich diversity of garments and body shapes and poses. We propose a framework for text-guided full-body human image synthesis via an attention-based latent code mapper, which enables more disentangled control of StyleGAN than existing mappers. Our latent code mapper adopts an attention mechanism that adaptively manipulates individual latent codes on different StyleGAN layers under text guidance. In addition, we introduce feature-space masking at inference time to avoid unwanted changes caused by text inputs. Our quantitative and qualitative evaluations reveal that our method can control generated images more faithfully to given texts than existing methods.