Search papers, labs, and topics across Lattice.
This paper introduces a novel dynamic mask-based backdoor attack against object detection models, using dataset poisoning to embed a malicious trigger. The attack leverages SAM (Segment Anything Model) to generate dynamic masks for trigger placement, enhancing stealthiness compared to static trigger methods. Experiments on a mushroom detection dataset using YOLOv7 demonstrate high accuracy on clean data alongside high attack success rates on poisoned samples, highlighting the vulnerability of outsourced AI models.
By dynamically masking triggers with SAM, this backdoor attack achieves stealth and success rates that surpass traditional static methods, even against robust object detectors like YOLOv7.
Deep learning has revolutionized numerous tasks within the computer vision field, including image classification, image segmentation, and object detection. However, the increasing deployment of deep learning models has exposed them to various adversarial attacks, including backdoor attacks. This paper presents a novel dynamic mask-based backdoor attack method, specifically designed for object detection models. We exploit a dataset poisoning technique to embed a malicious trigger, rendering any models trained on this compromised dataset vulnerable to our backdoor attack. We particularly focus on a mushroom detection dataset to demonstrate the practical risks posed by such attacks on critical real-life domains. Our work also emphasizes the importance of creating a detailed backdoor attack scenario to illustrate the significant risks associated with the outsourcing practice. Our approach leverages SAM, a recent and powerful image segmentation AI model, to create masks for dynamic trigger placement, introducing a new and stealthy attack method. Through extensive experimentation, we show that our sophisticated attack scenario maintains high accuracy on clean data with the YOLOv7 object detection model while achieving high attack success rates on poisoned samples. Our approach surpasses traditional methods for backdoor injection, which are based on static and consistent patterns. Our findings underscore the urgent need for robust countermeasures to protect deep learning models from these evolving adversarial threats.
