Contrastive Spectral Rectification: Test-Time Defense towards Zero-shot Adversarial Robustness of CLIP

Documenti analoghi

• V-Attack: Targeting Disentangled Value Features for Controllable Adversarial Attacks on LVLMs
  di: Nie, Sen, et al.
  Pubblicazione: (2025)
• What Makes VLMs Robust? Towards Reconciling Robustness and Accuracy in Vision-Language Models
  di: Nie, Sen, et al.
  Pubblicazione: (2026)
• CLIP is Strong Enough to Fight Back: Test-time Counterattacks towards Zero-shot Adversarial Robustness of CLIP
  di: Xing, Songlong, et al.
  Pubblicazione: (2025)
• Pre-trained Model Guided Fine-Tuning for Zero-Shot Adversarial Robustness
  di: Wang, Sibo, et al.
  Pubblicazione: (2024)
• FullLoRA: Efficiently Boosting the Robustness of Pretrained Vision Transformers
  di: Yuan, Zheng, et al.
  Pubblicazione: (2024)