WSMS: weakly supervised multi-feature steganalysis with EVT calibration for cloud-edge collaborative intelligence

Key Points

• Weakly supervised multi-feature steganalysis improves accuracy in cloud-edge scenarios while minimizing false positives.
• Key evidence shows a significant enhancement in detection rates, utilizing EVT calibration methods for optimization.
• Analysis of collaborative intelligence frameworks indicates that cloud-edge processing can enhance steganalysis efficiency.
• Results imply broader applications in cybersecurity, with further validation needed to support large-scale implementation.