Instruct2Attack: Language-Guided Semantic Adversarial Attacks

2023-11-27 · Jiang Liu, Chen Wei, Yuxiang Guo, Heng Yu, Alan Yuille, Soheil Feizi, Chun Pong Lau, Rama Chellappa

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.

📄 PDF Abstract BibTeX arXiv:2311.15551

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Attention 설명 없음

Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…

Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…

BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…

Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.

Exploring Semantic-constrained Adversarial Example with Instruction Uncertainty Reduction

2025-10-27 · Jin Hu, Jiakai Wang, Linna Jing, Haolin Li 외 arxiv

Recently, semantically constrained adversarial examples (SemanticAE), which are directly generated from natural language instructions, have become a promising avenue for future research due to their flexible attacking fo…

Visual Adversarial Attack on Vision-Language Models for Autonomous Driving

2024-11-27 · Tianyuan Zhang, Lu Wang, Xinwei Zhang, Yitong Zhang 외

Vision-language models (VLMs) have significantly advanced autonomous driving (AD) by enhancing reasoning capabilities. However, these models remain highly vulnerable to adversarial attacks. While existing research has pr…

Adversarial Attack Autonomous Driving Language Modelling Large Language Model

InstructTA: Instruction-Tuned Targeted Attack for Large Vision-Language Models

2023-12-04 · Xunguang Wang, Zhenlan Ji, Pingchuan Ma, Zongjie Li 외

Large vision-language models (LVLMs) have demonstrated their incredible capability in image understanding and response generation. However, this rich visual interaction also makes LVLMs vulnerable to adversarial examples…

Adversarial Attack Language Modelling Large Language Model Response Generation

AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red Teaming

2025-10-09 · Muxi Diao, Yutao Mou, Keqing He, Hanbo Song 외 arxiv

The safety of Large Language Models (LLMs) is crucial for the development of trustworthy AI applications. Existing red teaming methods often rely on seed instructions, which limits the semantic diversity of the synthesiz…

Red Teaming

When Alignment Fails: Multimodal Adversarial Attacks on Vision-Language-Action Models

2025-11-20 · Yuping Yan, Yuhan Xie, Yixin Zhang, Lingjuan Lyu 외 arxiv

Vision-Language-Action models (VLAs) have recently demonstrated remarkable progress in embodied environments, enabling robots to perceive, reason, and act through unified multimodal understanding. Despite their impressiv…

Semantic correspondence Adversarial Robustness