AI Paper Atlas

Research map / Image generation / Editing and style transfer

HyperReenact: One-Shot Reenactment via Jointly Learning to Refine and Retarget Faces

Stella Bounareli, Christos Tzelepis, Vasileios Argyriou, Ioannis Patras, Georgios Tzimiropoulos

ICCV 2023 · Image generation › Editing and style transfer · approach: Identity-preserving face manipulation · 51 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train with curriculum learning (inversion, then self, then cross-subject reenactment) (ablation)HyperReenact · VoxCeleb1 self reenactmentAED: 6.7 -> 5.1Improved
add curriculum learning (ablation)HyperReenact · VoxCeleb1 cross-subject reenactmentCSIM: 0.72 -> 0.68Hurt
add cross-subject reenactment training (ablation)HyperReenact · VoxCeleb1 cross-subject reenactmentCSIM: 0.53 -> 0.68Improved
add gaze loss (ablation)HyperReenact · VoxCeleb1 real image inversionGaze Error: 0.34 -> 0.24Improved
add gaze loss (ablation)HyperReenact · VoxCeleb1 cross-subject reenactmentGaze Error: 0.40 -> 0.31Improved
hypernetwork refinement plus pose retargeting vs prior methods on extreme pose pairsHyperReenact · VoxCeleb1 large head pose differencesCSIM: 0.53 -> 0.58 vs RomeImproved
use shared reenactment blocks for first seven layershypernetwork · model sizetrainable parameters: 1.2B -> 300MImproved
HyperReenact vs Rome on expression distance in cross-subject reenactmentHyperReenact · VoxCeleb1 cross-subject reenactmentAED: 9.3 vs 8.8 (Rome)No clear effect

Later papers that compared with this method

1 later accepted paper reports HyperReenact as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy