Instruct2Attack: Language-Guided Semantic Adversarial Attacks

TL;DR

Instruct2Attack (I2A) is a novel semantic adversarial attack method that uses language instructions and diffusion models to generate natural, diverse, and controllable adversarial examples capable of fooling advanced neural networks.

Contribution

The paper introduces I2A, a new language-guided semantic attack leveraging diffusion models, enhancing naturalness, diversity, and interpretability over previous noise-based methods.

Findings

I2A effectively breaks state-of-the-art neural networks.

I2A produces more natural and diverse adversarial examples.

The attack demonstrates strong transferability across architectures.

Abstract

We propose Instruct2Attack (I2A), a language-guided semantic attack that generates semantically meaningful perturbations according to free-form language instructions. We make use of state-of-the-art latent diffusion models, where we adversarially guide the reverse diffusion process to search for an adversarial latent code conditioned on the input image and text instruction. Compared to existing noise-based and semantic attacks, I2A generates more natural and diverse adversarial examples while providing better controllability and interpretability. We further automate the attack process with GPT-4 to generate diverse image-specific text instructions. We show that I2A can successfully break state-of-the-art deep neural networks even under strong adversarial defenses, and demonstrate great transferability among a variety of network architectures.