New text steganography uses dynamic codebook and multimodal LLM

Researchers have developed a novel black-box text steganography method that enhances security and practicality by employing a dynamic codebook and a multimodal large language model. This approach addresses limitations of existing white-box methods, which are prone to exposure, and black-box methods that lack flexibility due to fixed codebooks. The new technique constructs a dynamic codebook using shared session configurations and a multimodal LLM, embeds secret messages through an encrypted steganographic mapping, and utilizes a feedback optimization mechanism for accurate extraction. Experimental results indicate superior embedding capacity and text quality compared to existing white-box methods, along with improved practicality and flexibility over current black-box paradigms in social network environments. AI

IMPACT Enhances security and practicality for secret message embedding in digital communications.

RANK_REASON The cluster contains a research paper published on arXiv detailing a new technical method. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Alice
• alphaXiv
• arXiv
• Bob
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• Jianxin Gao
• large language models
• multimodal large language model
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →