4 papers
EVLM: Self-Reflective Multimodal Reasoning for Cross-Dimensional Visual Editing
Umar Khalid, Kashif Munir, Hasan Iqbal +6
Editing complex visual content from ambiguous or partially specified instructions remains a core challenge in vision-language modeling. Existing models can contextualize content bu…
On Transfer-based Universal Attacks in Pure Black-box Setting
Mohammad A. A. K. Jalwana, Naveed Akhtar, Ajmal Mian +2
Despite their impressive performance, deep visual models are susceptible to transferable black-box adversarial attacks. Principally, these attacks craft perturbations in a target m…
Fisher Information guided Purification against Backdoor Attacks
Nazmul Karim, Abdullah Al Arafat, Adnan Siraj Rakin +2
Studies on backdoor attacks in recent years suggest that an adversary can compromise the integrity of a deep neural network (DNN) by manipulating a small set of training samples. O…
Augmented Neural Fine-Tuning for Efficient Backdoor Purification
Nazmul Karim, Abdullah Al Arafat, Umar Khalid +2
Recent studies have revealed the vulnerability of deep neural networks (DNNs) to various backdoor attacks, where the behavior of DNNs can be compromised by utilizing certain types…