Trojvit: Trojan insertion in vision transformers
Abstract Vision Transformers (ViTs) have demonstrated the state-of-the-art performance in
various vision-related tasks. The success of ViTs motivates adversaries to perform backdoor …
various vision-related tasks. The success of ViTs motivates adversaries to perform backdoor …
A Comprehensive Survey on Backdoor Attacks and their Defenses in Face Recognition Systems
Deep learning has significantly transformed face recognition, enabling the deployment of
large-scale, state-of-the-art solutions worldwide. However, the widespread adoption of deep …
large-scale, state-of-the-art solutions worldwide. However, the widespread adoption of deep …
A dual stealthy backdoor: From both spatial and frequency perspectives
Backdoor attacks pose serious security threats to deep neural networks (DNNs).
Backdoored models make arbitrarily (targeted) incorrect predictions on inputs containing …
Backdoored models make arbitrarily (targeted) incorrect predictions on inputs containing …
MBA: Backdoor Attacks Against 3D Mesh Classifier
3D mesh classification deep neural network (3D DNN) has been widely applied in many
safety-critical domains. Backdoor attack is a serious threat that occurs during the training …
safety-critical domains. Backdoor attack is a serious threat that occurs during the training …
Look, listen, and attack: Backdoor attacks against video action recognition
Deep neural networks (DNNs) are vulnerable to a class of attacks called" backdoor attacks",
which create an association between a backdoor trigger and a target label the attacker is …
which create an association between a backdoor trigger and a target label the attacker is …
Look Listen and Attack: Backdoor Attacks Against Video Action Recognition
HA Al Kader Hammoud, S Liu… - Proceedings of the …, 2024 - openaccess.thecvf.com
Deep neural networks (DNNs) are vulnerable to a class of attacks called" backdoor attacks".
which create an association between a backdoor trigger and a target label the attacker is …
which create an association between a backdoor trigger and a target label the attacker is …
M-to-N Backdoor Paradigm: A Multi-Trigger and Multi-Target Attack to Deep Learning Models
Deep neural networks (DNNs) are vulnerable to backdoor attacks, where a backdoored
model behaves normally with clean inputs but exhibits attacker-specified behaviors upon the …
model behaves normally with clean inputs but exhibits attacker-specified behaviors upon the …
Untargeted backdoor attack against deep neural networks with imperceptible trigger
Recent research works have demonstrated that deep neural networks (DNNs) are
vulnerable to backdoor attacks. The existing backdoor attacks can only cause targeted …
vulnerable to backdoor attacks. The existing backdoor attacks can only cause targeted …
Attacks in adversarial machine learning: A systematic survey from the life-cycle perspective
Adversarial machine learning (AML) studies the adversarial phenomenon of machine
learning, which may make inconsistent or unexpected predictions with humans. Some …
learning, which may make inconsistent or unexpected predictions with humans. Some …
The Victim and The Beneficiary: Exploiting a Poisoned Model to Train a Clean Model on Poisoned Data
Z Zhu, R Wang, C Zou, L Jing - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Recently, backdoor attacks have posed a serious security threat to the training process of
deep neural networks (DNNs). The attacked model behaves normally on benign samples …
deep neural networks (DNNs). The attacked model behaves normally on benign samples …