INFAMOUS-NeRF: ImproviNg FAce MOdeling Using Semantically-Aligned Hypernetworks with Neural Radiance Fields

TL;DR

INFAMOUS-NeRF enhances face modeling by integrating hypernetworks with NeRF, achieving better representation, semantic alignment, and rendering quality without large pretrained models, through novel constraints and adaptive sampling.

Contribution

The paper introduces a hypernetwork-based NeRF for face modeling that improves representation and editability without large pretrained models, and proposes new constraints and sampling methods.

Findings

Outperforms prior face modeling methods in controlled settings.

Achieves higher representation power in in-the-wild scenarios.

Improves rendering quality near face boundaries.

Abstract

We propose INFAMOUS-NeRF, an implicit morphable face model that introduces hypernetworks to NeRF to improve the representation power in the presence of many training subjects. At the same time, INFAMOUS-NeRF resolves the classic hypernetwork tradeoff of representation power and editability by learning semantically-aligned latent spaces despite the subject-specific models, all without requiring a large pretrained model. INFAMOUS-NeRF further introduces a novel constraint to improve NeRF rendering along the face boundary. Our constraint can leverage photometric surface rendering and multi-view supervision to guide surface color prediction and improve rendering near the surface. Finally, we introduce a novel, loss-guided adaptive sampling method for more effective NeRF training by reducing the sampling redundancy. We show quantitatively and qualitatively that our method achieves higher representation power than prior face modeling methods in both controlled and in-the-wild settings. Code and models will be released upon publication.