GazeNeRF: 3D-Aware Gaze Redirection with Neural Radiance Fields

TL;DR

GazeNeRF introduces a 3D-aware neural radiance field approach for gaze redirection, enabling more accurate and consistent gaze control by modeling face and eyes as separate 3D structures.

Contribution

It presents a novel two-stream neural radiance field architecture that separately models face and eye regions for improved gaze redirection.

Findings

Outperforms previous 2D methods in accuracy and identity preservation.

Achieves more consistent 3D gaze redirection results.

Demonstrates effective control over gaze angles using 3D transformations.

Abstract

We propose GazeNeRF, a 3D-aware method for the task of gaze redirection. Existing gaze redirection methods operate on 2D images and struggle to generate 3D consistent results. Instead, we build on the intuition that the face region and eyeballs are separate 3D structures that move in a coordinated yet independent fashion. Our method leverages recent advancements in conditional image-based neural radiance fields and proposes a two-stream architecture that predicts volumetric features for the face and eye regions separately. Rigidly transforming the eye features via a 3D rotation matrix provides fine-grained control over the desired gaze angle. The final, redirected image is then attained via differentiable volume compositing. Our experiments show that this architecture outperforms naively conditioned NeRF baselines as well as previous state-of-the-art 2D gaze redirection methods in terms of redirection accuracy and identity preservation.