Lune

Profiles2022-2026

Yechao Zhang

Nanyang Technological University

12Papers
0Last author
28Unique co-authors
3First author
0Solo

First and last author come from byline position, which some venues assign alphabetically.

Recent papers12

  1. Dual-View Inference Attack: Machine Unlearning Amplifies Privacy ExposureAuthor 5 of 10AAAI 2026 · 3 citations
  2. VideoSEAL: Mitigating Evidence Misalignment in Agentic Long Video Understanding by Decoupling Answer AuthorityAuthor 2 of 5ICML 2026 · 1 citation
  3. MARS: A Malignity-Aware Backdoor Defense in Federated LearningAuthor 7 of 11NeurIPS 2025 · 16 citations
  4. Improving Generalization of Universal Adversarial Perturbation via Dynamic Maximin OptimizationAuthor 1 of 7AAAI 2025 · 7 citations
  5. Transferable Direct Prompt Injection via Activation-Guided MCMC SamplingAuthor 3 of 7EMNLP 2025 · 1 citation
  6. Secure Transfer Learning: Training Clean Model Against Backdoor in Pre-Trained Encoder and Downstream DatasetAuthor 1 of 7S&P 2025
  7. Why Does Little Robustness Help? A Further Step Towards Understanding Adversarial TransferabilityAuthor 1 of 9S&P 2024 · 36 citations
  8. Securely Fine-tuning Pre-trained Encoders Against Adversarial ExamplesAuthor 5 of 11S&P 2024 · 23 citations
  9. Unlearnable 3D Point Clouds: Class-wise Transformation Is All You NeedAuthor 6 of 8NeurIPS 2024 · 23 citations
  10. AdvCLIP: Downstream-agnostic Adversarial Examples in Multimodal Contrastive LearningAuthor 5 of 6ACM MM 2023 · 62 citations
  11. Protecting Facial Privacy: Generating Adversarial Identity Masks via Style-robust Makeup TransferAuthor 3 of 7CVPR 2022 · 123 citations
  12. BadHash: Invisible Backdoor Attacks against Deep Hashing with Clean LabelAuthor 3 of 4ACM MM 2022 · 36 citations
Dusk over the sea between two cliffs drawn in fine vertical lines