Input-Aware Dynamic Backdoor Attack

10/16/2020
by   Anh Nguyen, et al.
0

In recent years, neural backdoor attack has been considered to be a potential security threat to deep learning systems. Such systems, while achieving the state-of-the-art performance on clean data, perform abnormally on inputs with predefined triggers. Current backdoor techniques, however, rely on uniform trigger patterns, which are easily detected and mitigated by current defense methods. In this work, we propose a novel backdoor attack technique in which the triggers vary from input to input. To achieve this goal, we implement an input-aware trigger generator driven by diversity loss. A novel cross-trigger test is applied to enforce trigger nonreusablity, making backdoor verification impossible. Experiments show that our method is efficient in various attack scenarios as well as multiple datasets. We further demonstrate that our backdoor can bypass the state of the art defense methods. An analysis with a famous neural network inspector again proves the stealthiness of the proposed attack. Our code is publicly available at https://github.com/VinAIResearch/input-aware-backdoor-attack-release.

READ FULL TEXT

page 2

page 7

page 15

page 16

page 17

research
07/28/2023

Beating Backdoor Attack at Its Own Game

Deep neural networks (DNNs) are vulnerable to backdoor attack, which doe...
research
11/15/2022

Backdoor Attacks for Remote Sensing Data with Wavelet Transform

Recent years have witnessed the great success of deep learning algorithm...
research
06/14/2023

Your Email Address Holds the Key: Understanding the Connection Between Email and Password Security with Deep Learning

In this work, we investigate the effectiveness of deep-learning-based pa...
research
08/22/2022

An Input-Aware Mimic Defense Theory and its Practice

The current security problems in cyberspace are characterized by strong ...
research
12/20/2020

Color Channel Perturbation Attacks for Fooling Convolutional Neural Networks and A Defense Against Such Attacks

The Convolutional Neural Networks (CNNs) have emerged as a very powerful...
research
10/26/2021

Semantic Host-free Trojan Attack

In this paper, we propose a novel host-free Trojan attack with triggers ...
research
06/03/2021

Defending against Backdoor Attacks in Natural Language Generation

The frustratingly fragile nature of neural network models make current n...

Please sign up or login with your details

Forgot password? Click here to reset