HyperInverter: Improving StyleGAN Inversion via Hypernetwork

by   Tan M. Dinh, et al.

Real-world image manipulation has achieved fantastic progress in recent years as a result of the exploration and utilization of GAN latent spaces. GAN inversion is the first step in this pipeline, which aims to map the real image to the latent code faithfully. Unfortunately, the majority of existing GAN inversion methods fail to meet at least one of the three requirements listed below: high reconstruction quality, editability, and fast inference. We present a novel two-phase strategy in this research that fits all requirements at the same time. In the first phase, we train an encoder to map the input image to StyleGAN2 𝒲-space, which was proven to have excellent editability but lower reconstruction quality. In the second phase, we supplement the reconstruction ability in the initial phase by leveraging a series of hypernetworks to recover the missing information during inversion. These two steps complement each other to yield high reconstruction quality thanks to the hypernetwork branch and excellent editability due to the inversion done in the 𝒲-space. Our method is entirely encoder-based, resulting in extremely fast inference. Extensive experiments on two challenging datasets demonstrate the superiority of our method.


page 15

page 16

page 18

page 19

page 20

page 21

page 22

page 24


Cycle Encoding of a StyleGAN Encoder for Improved Reconstruction and Editability

GAN inversion aims to invert an input image into the latent space of a p...

Feature-Style Encoder for Style-Based GAN Inversion

We propose a novel architecture for GAN inversion, which we call Feature...

Photo-Realistic Out-of-domain GAN inversion via Invertibility Decomposition

The fidelity of Generative Adversarial Networks (GAN) inversion is imped...

ReGANIE: Rectifying GAN Inversion Errors for Accurate Real Image Editing

The StyleGAN family succeed in high-fidelity image generation and allow ...

Editing Out-of-domain GAN Inversion via Differential Activations

Despite the demonstrated editing capacity in the latent space of a pretr...

TriPlaneNet: An Encoder for EG3D Inversion

Recent progress in NeRF-based GANs has introduced a number of approaches...

Real-Time Radiance Fields for Single-Image Portrait View Synthesis

We present a one-shot method to infer and render a photorealistic 3D rep...

Please sign up or login with your details

Forgot password? Click here to reset