New framework enables covert communication by embedding data within semantic features

Researchers have developed an adaptive dual-path framework for covert semantic communication, integrating hidden message transmission with task-oriented semantic coding. This novel architecture embeds covert data within task-specific features using semantic-level intrinsic encoding, rather than traditional power-domain signal superposition. Experiments on the Cityscapes dataset showed the method achieved a near-random guessing level of 56.12% for an attacker while maintaining superior performance on primary semantic tasks. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a new method for secure data transmission within AI systems, potentially impacting privacy and security in semantic communication applications.

RANK_REASON This is a research paper published on arXiv detailing a novel technical framework. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Cityscapes

• paper
• other