Strategy to improve the accuracy of convolutional neural network architectures applied to digital image steganalysis in the spatial domain

Strategy to improve the accuracy of convolutional neural network architectures applied to digital image steganalysis in the spatial domain

In recent years, Deep Learning techniques applied to steganalysis have surpassed the traditional two-stage approach by unifying feature extraction and classification in a single model, the Convolutional Neural Network (CNN). Several CNN architectures have been proposed to solve this task, improving...

Full description

Bibliographic Details
Main Authors: Reinel Tabares-Soto, Harold Brayan Arteaga-Arteaga, Alejandro Mora-Rubio, Mario Alejandro Bravo-Ortíz, Daniel Arias-Garzón, Jesús Alejandro Alzate Grisales, Alejandro Burbano Jacome, Simon Orozco-Arias, Gustavo Isaza, Raul Ramos Pollan
Format: Article
Language:English
Published: PeerJ Inc. 2021-04-01
Series:PeerJ Computer Science
Subjects:
Convolutional neural network
Deep learning
Steganalysis
Strategy
Online Access:https://peerj.com/articles/cs-451.pdf

Internet

https://peerj.com/articles/cs-451.pdf

Similar Items