Hongbin Li

The paper introduces ImageProtector, a user-side method that embeds an imperceptible perturbation into images to prevent Multi-modal Large Language Models (MLLMs) from analyzing and extracting sensitive information from them.

This paper systematically studies the robustness of vision foundation models to common image perturbations, finding that most models are generally non-robust and proposing a fine-tuning method to improve this resilience.

This paper proposes a structured canonical polyadic decomposition (SCPD) algorithm for accurate sensing and joint network-level synchronization in networked integrated sensing and communication (ISAC), achieving superior accuracy and outlier robustness.