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

facebooktwittermailshare

UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS

Abstract: 

We propose a unified compression framework that uses generative adversarial networks (GAN) to compress image and speech signals. The compressed signal is represented by a latent vector fed into a generator network which is trained to produce high-quality signals that minimize a target objective function. To efficiently quantize the compressed signal, non-uniformly quantized optimal latent vectors are identified by iterative back-propagation with ADMM optimization performed for each iteration. Our experiments show that the proposed algorithm outperforms prior signal compression methods for both image and speech compression quantified in various metrics including bit rate, PSNR, and neural network-based signal classification accuracy.

up
0 users have voted:

Paper Details

Authors:
Submitted On:
16 May 2020 - 4:30pm
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Bowen Liu
Paper Code:
4405
Document Year:
2020
Cite

Document Files

4405.pdf

(16)

Subscribe

[1] , "UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS", IEEE SigPort, 2020. [Online]. Available: http://sigport.org/5379. Accessed: Jul. 14, 2020.
@article{5379-20,
url = {http://sigport.org/5379},
author = { },
publisher = {IEEE SigPort},
title = {UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS},
year = {2020} }
TY - EJOUR
T1 - UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS
AU -
PY - 2020
PB - IEEE SigPort
UR - http://sigport.org/5379
ER -
. (2020). UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS. IEEE SigPort. http://sigport.org/5379
, 2020. UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS. Available at: http://sigport.org/5379.
. (2020). "UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS." Web.
1. . UNIFIED SIGNAL COMPRESSION USING GENERATIVE ADVERSARIAL NETWORKS [Internet]. IEEE SigPort; 2020. Available from : http://sigport.org/5379