Security
I2I Backdoor: Backdoor Attacks against Image-to-Image Tasks
arXiv cs.CR0 views
arXiv:2407.10445v2 Announce Type: replace-cross Abstract: With the rapid development of deep learning technology, deep learning-based Image-to-Image (I2I) networks have become the predominant choice for I2I tasks like image super-resolution and denoising. Despite their remarkable performance, the security of I2I networks has not been thoroughly investigated. While some studies have probed their susceptibility to adversarial attacks, none have explored the backdoor attack against I2I networks, which is a more stealthy and severe threat. In this work, for the first time, we comprehensively investigate the vulnerability of I2I networks to backdoor attacks. We propose a backdoor attack against I2I tasks, where the backdoored I2I network behaves normally on clean input images, yet outputs a specific inappropriate image when the backdoor trigger appears on the input image. To achieve such an I2I backdoor attack, we design a universal adversarial perturbation (UAP) generation algorithm for I2I networks, where the generated UAP is used as the trigger for the I2I backdoor. Besides, multi-task learning (MTL) with dynamic weighting methods is employed in the backdoor training process to gain better results. Expanding our focus beyond I2I tasks, we extend our I2I backdoor to attack downstream tasks, including image classification and object detection. Specifically, the backdoor-triggered image processed by the backdoored image denoising network can fool the downstream image classifiers and object detectors. Extensive experiments demonstrate the effectiveness of the I2I backdoor on state-of-the-art I2I network architectures as well as the robustness against different backdoor defenses.
Source:https://arxiv.org/abs/2407.10445PreviousAdversarial Process Disturbances at a Deployable Alarm Budget: Detector Evaluation on a Simulator and a Real PlantNextVerifyPrompt: Towards Text-to-Image Models Verification Behind Black-Box APIs
Related posts
Analysis of Commit Signing on GitHub
0arXiv cs.CR
Post-Quantum Cryptography Education for Software Developers: A Multivocal Literature Review
0arXiv cs.CR
Beyond Stars: MTTX Brings Developer Responsiveness to Dependency Decisions, from Dashboard to IDE to GitHub Actions
0arXiv cs.CR
ChainIT Transaction Truth: Stablecoins, Web3 Wallets, and Multi-Rail Commerce. A Zero-Trust Control Plane for Verified Identity, Organizational Authority, Runtime Compliance, Secure Wallet Execution,
0arXiv cs.CR
CVE-2026-71861: CVE-2026-71861: Authentication Bypass via Unsigned Binding Headers in Chainguard Go SDK
0dev.to Security
Google's Gemini 4 Neon Rivals Top AI Models At Half The Cost
0dev.to Machine Learning