ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper FaceBERT: Face De-Identification Using VQGAN and BERT
Cited 3 time in scopus Share share facebook twitter linkedin kakaostory
Authors
Dong-Hyuck Im, Yong-Seok Seo
Issue Date
2022-10
Citation
International Conference on Information and Communication Technology Convergence (ICTC) 2022, pp.2013-2015
Publisher
IEEE
Language
English
Type
Conference Paper
DOI
https://dx.doi.org/10.1109/ICTC55196.2022.9952883
Abstract
This paper presents FaceBERT, a face de-identification technique based on image generation to solve the infringement of portrait rights. Using the VQGAN model, it learns a quantized codebook that expresses an image in block units, encodes the image using the codebook, and then trains the BERT model. As a result of an experiment using FFHQ and CelebA-HQ, it was confirmed that the face images generated using FaceBERT were natural de-identified face images and different from the originals.
KSP Keywords
Face image, Identification technique, face de-identification, image generation