AgeSynth - Latent Age Direction

Idiap Research Institute - Main Repository 2026 Dataset Open access

Read full text
Authors: Luis Santiago Luevano Garcia, Pavel Korshunov, Sébastien Marcel
DOI: 10.34777/z9dr-sr90
Cited by: 0

Abstract

A 512-D latent direction (normal vector) in StyleGAN2’s $\mathcal{W}$ space that enables smooth, age manipulation of human faces. The vector was obtained via Support-Vector Regression (SVR) trained to predict chronological age from latent codes projected from UTKFace images. Traversing the vector produces synthetic age progression/regression. This latent direction was computed on the official StyleGAN2 1024x1024 model.

Fields of Study & Subjects

Generative Adversarial Networks and Image Synthesis Face recognition and analysis Digital Media Forensic Detection

Publication details

Journal / source
Idiap Research Institute - Main Repository
Publication date
19 January 2026
Language
EN
Document type
Dataset