TERD: A Unified Framework for Safeguarding Diffusion Models Against Backdoors
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910595145531392 |
|---|---|
| author | Mo, Yichuan Huang, Hui Li, Mingjie Li, Ang Wang, Yisen |
| author_facet | Mo, Yichuan Huang, Hui Li, Mingjie Li, Ang Wang, Yisen |
| contents | Diffusion models have achieved notable success in image generation, but they remain highly vulnerable to backdoor attacks, which compromise their integrity by producing specific undesirable outputs when presented with a pre-defined trigger. In this paper, we investigate how to protect diffusion models from this dangerous threat. Specifically, we propose TERD, a backdoor defense framework that builds unified modeling for current attacks, which enables us to derive an accessible reversed loss. A trigger reversion strategy is further employed: an initial approximation of the trigger through noise sampled from a prior distribution, followed by refinement through differential multi-step samplers. Additionally, with the reversed trigger, we propose backdoor detection from the noise space, introducing the first backdoor input detection approach for diffusion models and a novel model detection algorithm that calculates the KL divergence between reversed and benign distributions. Extensive evaluations demonstrate that TERD secures a 100% True Positive Rate (TPR) and True Negative Rate (TNR) across datasets of varying resolutions. TERD also demonstrates nice adaptability to other Stochastic Differential Equation (SDE)-based models. Our code is available at https://github.com/PKU-ML/TERD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_05294 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | TERD: A Unified Framework for Safeguarding Diffusion Models Against Backdoors Mo, Yichuan Huang, Hui Li, Mingjie Li, Ang Wang, Yisen Cryptography and Security Artificial Intelligence Machine Learning Diffusion models have achieved notable success in image generation, but they remain highly vulnerable to backdoor attacks, which compromise their integrity by producing specific undesirable outputs when presented with a pre-defined trigger. In this paper, we investigate how to protect diffusion models from this dangerous threat. Specifically, we propose TERD, a backdoor defense framework that builds unified modeling for current attacks, which enables us to derive an accessible reversed loss. A trigger reversion strategy is further employed: an initial approximation of the trigger through noise sampled from a prior distribution, followed by refinement through differential multi-step samplers. Additionally, with the reversed trigger, we propose backdoor detection from the noise space, introducing the first backdoor input detection approach for diffusion models and a novel model detection algorithm that calculates the KL divergence between reversed and benign distributions. Extensive evaluations demonstrate that TERD secures a 100% True Positive Rate (TPR) and True Negative Rate (TNR) across datasets of varying resolutions. TERD also demonstrates nice adaptability to other Stochastic Differential Equation (SDE)-based models. Our code is available at https://github.com/PKU-ML/TERD. |
| title | TERD: A Unified Framework for Safeguarding Diffusion Models Against Backdoors |
| topic | Cryptography and Security Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2409.05294 |