Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES

Abstract: 

We present a novel variational generative adversarial network (VGAN) based on Wasserstein loss to learn a latent representation
from a face image that is invariant to identity but preserves head-pose information. This facilitates synthesis of a realistic face
image with the same head pose as a given input image, but with a different identity. One application of this network is in
privacy-sensitive scenarios; after identity replacement in an image, utility, such as head pose, can still
be recovered. Extensive experimental validation on synthetic and real human-face image datasets performed under 3 threat
scenarios confirms the ability of the proposed network to preserve head pose of the input image, mask the input identity,
and synthesize a good-quality realistic face image of a desired identity. We also show that our network can be used to perform

pose-preserving identity morphing and identity-preserving pose morphing. The proposed method improves over a recent state-
of-the-art method in terms of quantitative metrics as well as synthesized image quality.

up
0 users have voted:

Paper Details

Authors:
Jiawei Chen, Janusz Konrad, Prakash Ishwar
Submitted On:
11 October 2019 - 4:41pm
Short Link:
Type:
Poster
Event:
Presenter's Name:
Hiroki Kawai
Paper Code:
56
Document Year:
2019
Cite

Document Files

MLSP poster presentation

(16)

Subscribe

[1] Jiawei Chen, Janusz Konrad, Prakash Ishwar, "VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4856. Accessed: Nov. 19, 2019.
@article{4856-19,
url = {http://sigport.org/4856},
author = {Jiawei Chen; Janusz Konrad; Prakash Ishwar },
publisher = {IEEE SigPort},
title = {VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES},
year = {2019} }
TY - EJOUR
T1 - VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES
AU - Jiawei Chen; Janusz Konrad; Prakash Ishwar
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4856
ER -
Jiawei Chen, Janusz Konrad, Prakash Ishwar. (2019). VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES. IEEE SigPort. http://sigport.org/4856
Jiawei Chen, Janusz Konrad, Prakash Ishwar, 2019. VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES. Available at: http://sigport.org/4856.
Jiawei Chen, Janusz Konrad, Prakash Ishwar. (2019). "VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES." Web.
1. Jiawei Chen, Janusz Konrad, Prakash Ishwar. VAE/WGAN-BASED IMAGE REPRESENTATION LEARNING FOR POSE-PRESERVING SEAMLESS IDENTITY REPLACEMENT IN FACIAL IMAGES [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4856