Software is continually increasing in size and complexity, and therefore, vulnerability discovery would benefit from techniques that identify potentially vulnerable regions within large code bases, as this allows for easing vulnerability detection by reducing the search space. Previous work has explored the use of conventional code-quality and complexity metrics in highlighting suspicious sections of (source) code. Recently, researchers also proposed to reduce the vulnerability search space by studying code properties with neural networks. However, previous work generally failed in leveraging the rich metadata that is available for long-running, large code repositories.

A recent source of concern for the security of neural networks is the emergence of clean-label dataset poisoning attacks, wherein correctly labeled poison samples are injected into the training dataset. While these poison samples look legitimate to the human observer, they contain malicious characteristics that trigger a targeted misclassification during inference. We propose a scalable and transferable clean-label poisoning attack against transfer learning, which creates poison images with their center close to the target image in the feature space. Our attack, Bullseye Polytope, improves the attack success rate of the current state-of-the-art by 26.75% in end-to-end transfer learning, while increasing attack speed by a factor of 12. We further extend Bullseye Polytope to a more practical attack model by including multiple images of the same object (e.g., from different angles) when crafting the poison samples. We demonstrate that this extension improves attack transferability by over 16% to unseen images (of the same object) without using extra poison samples.

Deep learning has shown impressive performance on hard perceptual problems. However, researchers found deep learning systems to be vulnerable to small, specially crafted perturbations that are imperceptible to humans. Such perturbations cause deep learning systems to mis-classify adversarial examples, with potentially disastrous consequences where safety or security is crucial. Prior defenses against adversarial examples either targeted specific attacks or were shown to be ineffective.