Revisiting Latent Space of GAN Inversion for Real Image Editing
share
The exploration of the latent space in StyleGANs and GAN inversion exemplify impressive real-world image editing, yet the trade-off between reconstruction quality and editing quality remains an open problem. In this study, we revisit StyleGANs' hyperspherical prior 𝒵 and combine it with highly capable latent spaces to build combined spaces that faithfully invert real images while maintaining the quality of edited images. More specifically, we propose ℱ/𝒵^+ space consisting of two subspaces: ℱ space of an intermediate feature map of StyleGANs enabling faithful reconstruction and 𝒵^+ space of an extended StyleGAN prior supporting high editing quality. We project the real images into the proposed space to obtain the inverted codes, by which we then move along 𝒵^+, enabling semantic editing without sacrificing image quality. Comprehensive experiments show that 𝒵^+ can replace the most commonly-used 𝒲, 𝒲^+, and 𝒮 spaces while preserving reconstruction quality, resulting in reduced distortion of edited images.
READ FULL TEXT