
AUEditNet achieves accurate manipulation of facial action unit intensities in high-resolution synthetic face images, without requiring retraining or extra estimators, by leveraging a dual-branch architecture that implicitly disentangles facial attributes and identity even with limited subject data.


coremsg

accurate-facial-action-unit-intensity-manipulation-with-implicit-disentanglement-in-limited-subject-datasets


Accurate Facial Action Unit Intensity Manipulation with Implicit Disentanglement in Limited-Subject Datasets


title_rewrite


DiffFAE introduces a one-stage diffusion-based framework for high-fidelity facial appearance editing, addressing challenges of low generation fidelity, poor attribute preservation, and inefficient inference.


difffae-advancing-high-fidelity-one-shot-facial-appearance-editing


DiffFAE: Advancing High-fidelity One-shot Facial Appearance Editing
