B-AVIBench: Towards Evaluating the Robustness of Large Vision-Language Model on Black-box Adversarial Visual-Instructions

TL;DR

B-AVIBench is a comprehensive framework for evaluating the robustness of large vision-language models against various black-box adversarial visual-instructions and content biases, revealing vulnerabilities and biases in current models.

Contribution

Introduces B-AVIBench, a new benchmark and toolkit for assessing LVLM robustness to diverse adversarial and biased visual-instructions.

Findings

LVLMs are vulnerable to black-box adversarial visual-instructions.

Advanced models like GPT-4V exhibit inherent biases.

Extensive evaluation reveals critical robustness and bias issues.

Abstract

Large Vision-Language Models (LVLMs) have shown significant progress in responding well to visual-instructions from users. However, these instructions, encompassing images and text, are susceptible to both intentional and inadvertent attacks. Despite the critical importance of LVLMs' robustness against such threats, current research in this area remains limited. To bridge this gap, we introduce B-AVIBench, a framework designed to analyze the robustness of LVLMs when facing various Black-box Adversarial Visual-Instructions (B-AVIs), including four types of image-based B-AVIs, ten types of text-based B-AVIs, and nine types of content bias B-AVIs (such as gender, violence, cultural, and racial biases, among others). We generate 316K B-AVIs encompassing five categories of multimodal capabilities (ten tasks) and content bias. We then conduct a comprehensive evaluation involving 14 open-source LVLMs to assess their performance. B-AVIBench also serves as a convenient tool for practitioners to evaluate the robustness of LVLMs against B-AVIs. Our findings and extensive experimental results shed light on the vulnerabilities of LVLMs, and highlight that inherent biases exist even in advanced closed-source LVLMs like GeminiProVision and GPT-4V. This underscores the importance of enhancing the robustness, security, and fairness of LVLMs. The source code and benchmark are available at https://github.com/zhanghao5201/B-AVIBench.