Text-Aware, Vision-Blind: How Multimodal Image Editing Systems Handle Unsafe Inputs
[Kryptografia i bezpieczeństwo] Autorzy: Ziqing Yang, Yixin Wu, Michael Backes i 1 więcej
Multimodal image generation systems have made rapid progress and increasingly emphasize content safety. However, their safety mechanisms tend to rely on a single modality and lack joint consideration of both visual content and text prompts, particularly in instruction-guided image editing, where the uploaded image may itself carry risk. We thus introduce EditSafetyBench, a benchmark that measures system-level safety awareness of advanced image editing systems, containing 5,940 image-instruction pairs, covering safe and unsafe images across 11 risk categories and three edit intents. We then probe the benchmark dataset on four systems, including Gemini, GPT, Qwen API, and Qwen Local. Based on our evaluation, we observe a systematic text-vision safety mismatch: Gemini and GPT refuse 87.1\% and 81.6\% cases containing unsafe-intent editing prompts, while only refusing 24.2\% and 22.4\% case…
- Kategoria arxiv
- cs.CR
- Kategoria
- Kryptografia i bezpieczeństwo