This study introduces a new method for initiating a backdoor attack during runtime. Our approach uses feedback-based iterative optimization trigger method with an adaptive trigger generation to generate a trigger that guides the attack. Using a custom-weighted cross-entropy loss (objective function) makes our attack effective, which creates a bias when the trigger is added to the input. We conducted thorough experiments to demonstrate the versatility and strength of our approach on standard benchmarks, which enhances our understanding of backdoor attack dynamics. Our method advances the field of backdoor techniques and exposes new vulnerabilities in deep learning models. Researchers hoping to develop defensive strategies against sophisticated backdoor attacks will find our research to be a valuable reference point.

Deep neural networks are susceptible to various backdoor attacks, such as training time attacks, where the attacker can inject a trigger pattern into a small portion of the dataset to control the model's predictions at runtime. Backdoor attacks are dangerous because they do not degrade the model's performance. This article explores the feasibility of a new type of backdoor attack, a data-free backdoor. Unlike traditional backdoor attacks that require poisoning data and injection during training, our approach, the iterative optimization trigger method (IOTM), enables trigger generation without compromising the integrity of the models and datasets. We propose an attack based on an IOTM technique, guided by an adaptive trigger generator (ATG) and employing a custom objective function. ATG dynamically refines the trigger using feedback from the model's predictions. We empirically evaluated the effectiveness of IOTM with three deep learning models (CNN, VGG16, and ResNet18) using the CIFAR1...