Comparison of neural networks for suppression of multiplicative noise in images
The paper compares several neural network (NN) architectures for suppression of multiplicative noise. The images may contain sharp boundaries and large homogeneous areas. Convolutional and fully connected networks are investigated. It is shown that different architectures require significantly diffe...
Saved in:
Published in | Kompʹûternaâ optika Vol. 48; no. 3; pp. 425 - 431 |
---|---|
Main Authors | , , , , |
Format | Journal Article |
Language | English |
Published |
Samara National Research University
01.06.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Be the first to leave a comment!