{"id":"W7123736317","doi":"10.1109/mmsp64401.2025.11324270","title":"Secure INN-based Steganography via Model Smoothing and Adversarial Attacks","year":2025,"lang":"","type":"article","venue":"","topic":"Advanced Steganography and Watermarking Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Natural Science Foundation of Guangdong Province","keywords":"Steganography; Smoothing; Embedding; Adversarial system; Deep learning; Noise (video); Autoencoder; Pattern recognition (psychology); Cryptography","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000617258,0.0005354244,0.000642736,0.0002719196,0.0002562658,0.00033176,0.0005737806,0.0006305968,0.0006715789],"category_scores_gemma":[0.001829052,0.0002372231,0.0004461372,0.0002303841,0.0007170526,0.0009821521,0.001106289,0.000961397,0.0002165688],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003978625,"about_ca_system_score_gemma":0.000465587,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001017017,"about_ca_topic_score_gemma":0.00125177,"domain_scores_codex":[0.9996009,0.00009159023,0.00002079648,0.00008935156,0.00013765,0.00005963688],"domain_scores_gemma":[0.999283,0.0003560256,0.0001047091,0.0001531465,0.00007489273,0.00002817671],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002824738,0.00006970905,0.001340229,0.0001096187,0.00006856159,0.0002091779,0.0001441385,0.7659987,0.05020291,0.0282291,0.001100161,0.1522452],"study_design_scores_gemma":[0.000003499511,0.00002215669,0.00009432299,0.000003279451,0.000003522904,0.00004830806,0.000004710248,0.9895153,0.006967356,0.003026774,0.0003063331,0.000004425796],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05485597,0.0002122267,0.9425968,0.0001685964,0.00002433324,0.00001752912,0.00002665339,0.0006111633,0.001486754],"genre_scores_gemma":[0.9025821,0.000220436,0.09470351,0.00009111039,0.00002143182,0.00003318539,0.00007677036,0.00005661593,0.002214878],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.001017017,"threshold_uncertainty_score":0.003264368,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009918821075704784,"score_gpt":0.2577755558925942,"score_spread":0.2478567348168894,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}