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

facebooktwittermailshare

Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising

Abstract: 

While convolutional neural nets (CNN) have achieved remarkable performance for a wide range of inverse imaging applications, the filter coefficients are computed in a purely data-driven manner and are not explainable. Inspired by an analytically derived CNN by Hadji et al., in this paper we construct a new layered graph convolutional neural net (GCNN) using GraphBio as our graph filter. Unlike convolutional filters in previous GNNs, our employed GraphBio is analytically defined and requires no training, and we optimize the end-to-end system only via learning of appropriate graph topology at each layer. In signal filtering terms, it means that our linear graph filter at each layer is always intrepretable as low-pass with known biorthogonal conditions, while the graph spectrum itself is optimized via data training. As an example application, we show that our analytical GCNN achieves image denoising performance comparable to a state-of-the-art CNN-based scheme when the training and testing data share the same statistics, and when they differ, our analytical GCNN outperforms it by more than 1dB in PSNR.

up
0 users have voted:

Paper Details

Authors:
, Gene Cheung, Richard Wildes and Chia-Wen Lin
Submitted On:
15 May 2020 - 8:51pm
Short Link:
Type:
Presentation Slides
Event:
Document Year:
2020
Cite

Document Files

Icassp_DeepAGF_2020_slides

(69)

Subscribe

[1] , Gene Cheung, Richard Wildes and Chia-Wen Lin, "Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising", IEEE SigPort, 2020. [Online]. Available: http://sigport.org/5361. Accessed: Sep. 26, 2020.
@article{5361-20,
url = {http://sigport.org/5361},
author = {; Gene Cheung; Richard Wildes and Chia-Wen Lin },
publisher = {IEEE SigPort},
title = {Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising},
year = {2020} }
TY - EJOUR
T1 - Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising
AU - ; Gene Cheung; Richard Wildes and Chia-Wen Lin
PY - 2020
PB - IEEE SigPort
UR - http://sigport.org/5361
ER -
, Gene Cheung, Richard Wildes and Chia-Wen Lin. (2020). Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising. IEEE SigPort. http://sigport.org/5361
, Gene Cheung, Richard Wildes and Chia-Wen Lin, 2020. Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising. Available at: http://sigport.org/5361.
, Gene Cheung, Richard Wildes and Chia-Wen Lin. (2020). "Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising." Web.
1. , Gene Cheung, Richard Wildes and Chia-Wen Lin. Graph Neural Net using Analytical Graph Filters and Topology Optimization for Image Denoising [Internet]. IEEE SigPort; 2020. Available from : http://sigport.org/5361