Research
impact 16
Detecting Malicious Concepts without Image Generation in AI-Generated Content (AIGC)
Detecting Malicious Concepts without Image Generation in AI-Generated Content (AIGC) arXiv:2502.08921v2 Announce Type: replace Abstract: The task of text-to-image generation has achieved tremendous success in practice,…
Why it matters
This adds a new dimension to the generation conversation. Practitioners should assess exposure to detecting changes.