A Robust Classification-autoencoder to Defend Outliers and Adversaries

06/30/2021
by   Lijia Yu, et al.
17

In this paper, we present a robust classification-autoencoder (CAE) which has strong ability to recognize outliers and defend adversaries. The basic idea is to change the autoencoder from an unsupervised learning method into a classifier. The CAE is a modified autoencoder, where the encoder is used to compress samples with different labels into disjoint compression spaces and the decoder is used to recover a sample with a given label from the corresponding compression space. The encoder is used as a classifier and the decoder is used to decide whether the classification given by the encoder is correct by comparing the input sample with the output. Since adversary samples are seeming inevitable for the current DNN framework, we introduce the list classification based on CAE to defend adversaries, which outputs several labels and the corresponding samples recovered by the CAE. The CAE is evaluated using the MNIST dataset in great detail. It is shown that the CAE network can recognize almost all outliers and the list classification contains the correct label for almost all adversaries.

READ FULL TEXT

page 8

page 10

page 12

page 13

page 14

page 16

page 17

page 20

research
04/05/2022

LatentGAN Autoencoder: Learning Disentangled Latent Distribution

In autoencoder, the encoder generally approximates the latent distributi...
research
11/07/2020

Identifying Mislabeled Images in Supervised Learning Utilizing Autoencoder

Supervised learning is based on the assumption that the ground truth in ...
research
06/23/2022

On making optimal transport robust to all outliers

Optimal transport (OT) is known to be sensitive against outliers because...
research
08/10/2023

Symmetry Defense Against XGBoost Adversarial Perturbation Attacks

We examine whether symmetry can be used to defend tree-based ensemble cl...
research
10/26/2021

Improving Local Effectiveness for Global robust training

Despite its popularity, deep neural networks are easily fooled. To allev...
research
07/12/2020

Disentangled Variational Autoencoder based Multi-Label Classification with Covariance-Aware Multivariate Probit Model

Multi-label classification is the challenging task of predicting the pre...
research
06/03/2022

Is an encoder within reach?

The encoder network of an autoencoder is an approximation of the nearest...

Please sign up or login with your details

Forgot password? Click here to reset