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

facebooktwittermailshare

Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images

Abstract: 

Based on recent random matrix advances in the analysis of kernel methods for classification and clustering, this paper proposes the study of large kernel methods for a wide class of random inputs, i.e., concentrated data, which are more generic than Gaussian mixtures. The concentration assumption is motivated by the fact that one can use generative models to design complex data structures, through Lipschitz-ally transformed concentrated vectors (e.g., Gaussian) which remain concentrated vectors. Applied to spectral clustering, we demonstrate that our theoretical findings closely match the behavior of large kernel matrices, when considering the fed-in data as CNN representations of GAN-generated images (i.e., concentrated vectors by design).

up
0 users have voted:

Paper Details

Authors:
Mohamed El Amine Seddik, Mohamed Tamaazousti, Romain Couillet
Submitted On:
10 May 2019 - 4:02pm
Short Link:
Type:
Poster
Event:
Presenter's Name:
Mohamed El Amine Seddik
Paper Code:
2259
Document Year:
2019
Cite

Document Files

poster_icassp2019.pdf

(30)

Keywords

Additional Categories

Subscribe

[1] Mohamed El Amine Seddik, Mohamed Tamaazousti, Romain Couillet, "Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4398. Accessed: Aug. 25, 2019.
@article{4398-19,
url = {http://sigport.org/4398},
author = {Mohamed El Amine Seddik; Mohamed Tamaazousti; Romain Couillet },
publisher = {IEEE SigPort},
title = {Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images},
year = {2019} }
TY - EJOUR
T1 - Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images
AU - Mohamed El Amine Seddik; Mohamed Tamaazousti; Romain Couillet
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4398
ER -
Mohamed El Amine Seddik, Mohamed Tamaazousti, Romain Couillet. (2019). Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images. IEEE SigPort. http://sigport.org/4398
Mohamed El Amine Seddik, Mohamed Tamaazousti, Romain Couillet, 2019. Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images. Available at: http://sigport.org/4398.
Mohamed El Amine Seddik, Mohamed Tamaazousti, Romain Couillet. (2019). "Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images." Web.
1. Mohamed El Amine Seddik, Mohamed Tamaazousti, Romain Couillet. Kernel Random Matrices of Large Concentrated Data: The Example of GAN-Generated Images [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4398