Picture for Jiawei Liang

Jiawei Liang

Efficient Backdoor Defense in Multimodal Contrastive Learning: A Token-Level Unlearning Method for Mitigating Threats

Add code
Sep 29, 2024
Figure 1 for Efficient Backdoor Defense in Multimodal Contrastive Learning: A Token-Level Unlearning Method for Mitigating Threats
Figure 2 for Efficient Backdoor Defense in Multimodal Contrastive Learning: A Token-Level Unlearning Method for Mitigating Threats
Figure 3 for Efficient Backdoor Defense in Multimodal Contrastive Learning: A Token-Level Unlearning Method for Mitigating Threats
Figure 4 for Efficient Backdoor Defense in Multimodal Contrastive Learning: A Token-Level Unlearning Method for Mitigating Threats
Viaarxiv icon

Adversarial Backdoor Defense in CLIP

Add code
Sep 24, 2024
Viaarxiv icon

Revisiting Backdoor Attacks against Large Vision-Language Models

Add code
Jun 27, 2024
Viaarxiv icon

Unlearning Backdoor Threats: Enhancing Backdoor Defense in Multimodal Contrastive Learning via Local Token Unlearning

Add code
Mar 24, 2024
Viaarxiv icon

VL-Trojan: Multimodal Instruction Backdoor Attacks against Autoregressive Visual Language Models

Add code
Feb 21, 2024
Viaarxiv icon

Poisoned Forgery Face: Towards Backdoor Attacks on Face Forgery Detection

Add code
Feb 18, 2024
Viaarxiv icon

Rethinking Data Augmentation in Knowledge Distillation for Object Detection

Add code
Sep 20, 2022
Figure 1 for Rethinking Data Augmentation in Knowledge Distillation for Object Detection
Figure 2 for Rethinking Data Augmentation in Knowledge Distillation for Object Detection
Figure 3 for Rethinking Data Augmentation in Knowledge Distillation for Object Detection
Figure 4 for Rethinking Data Augmentation in Knowledge Distillation for Object Detection
Viaarxiv icon