Synthesized Dataset of Stylized and Real Face Pairs

A large-scale synthesized dataset of stylized face (SF) and ground-truth real face (RF) pairs is generated to train the Identity-preserving Face Recovery from Portraits (IFRP) method. This dataset includes 30K training pairs created from aligned real faces and various stylized representations.

Data and Resources

Cite this as

Fatemeh Shiri, Xin Yu, Fatih Porikli, Richard Hartley, Piotr Koniusz (2024). Dataset: Synthesized Dataset of Stylized and Real Face Pairs. https://doi.org/10.57702/2qg3nt0h

DOI retrieved: November 25, 2024

Additional Info

Field Value
Created November 25, 2024
Last update November 25, 2024
Defined In https://doi.org/10.48550/arXiv.1801.02279
Author Fatemeh Shiri
More Authors
Xin Yu
Fatih Porikli
Richard Hartley
Piotr Koniusz