About this document
3D Face Reconstruction from Templates by Dewi Yuliani is a document available to read on EtoBox.
The paper presents a novel template inversion attack method called GaFaR, which reconstructs 3D faces from facial templates used in face recognition systems. Utilizing a geometry-aware generator network based on generative neural radiance fields (GNeRF), the method combines real and synthetic data for training and optimizes camera parameters to enhance attack success rates. This research is significant as it marks the first instance of 3D face reconstruction from facial templates, addressing both security a
- Author
- Dewi Yuliani
- Language
- EN