← Search

Tenghui Li

2 accepted papers

2026

Two Modalities Are Better Than One: Efficient Adversarial Purification via Multimodal Diffusion Models

ICML 2026poster

Adversarial purification uses generative models to restore clean data distributions from unseen attacks without retraining classifiers. However, unimodal diffusion-based approaches struggle to preserve semantic consistency, while recent multimodal variants rely on computationally expensive adversari…

Cited by 0SourceScholar
2024

Diffusion Models Demand Contrastive Guidance for Adversarial Purification to Advance

ICML 2024poster

In adversarial defense, adversarial purification can be viewed as a special generation task with the purpose to remove adversarial attacks and diffusion models excel in adversarial purification for their strong generative power. With different predetermined generation requirements, various types of…

Cited by 6SourcePDFScholar