{"id":"W4393156222","doi":"10.1609/aaai.v38i7.28571","title":"A New Benchmark and Model for Challenging Image Manipulation Detection","year":2024,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Image Processing Techniques and Applications","field":"Engineering","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Benchmark (surveying); Artificial intelligence; Computer science; Image (mathematics); Computer vision; Image manipulation; Pattern recognition (psychology); Cartography; Geography","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001266496,0.0001128012,0.00009162849,0.00007931513,0.00009957694,0.0001759174,0.0001676803,0.00005511213,0.000009134938],"category_scores_gemma":[0.00003552802,0.00009485272,0.00004470963,0.000181239,0.00004112754,0.0002199468,0.0000320738,0.0001293503,0.000004289483],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002812085,"about_ca_system_score_gemma":0.00001764702,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006448739,"about_ca_topic_score_gemma":0.000004463149,"domain_scores_codex":[0.9993809,9.186205e-7,0.0002046955,0.0001836087,0.00009502094,0.0001348018],"domain_scores_gemma":[0.9997256,0.00002447275,0.00003540049,0.00006597723,0.0001150931,0.00003347445],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001181721,0.000008950017,0.000001522082,0.0003238484,0.000009167798,3.698072e-8,0.0006785443,0.001429142,0.3256306,0.3275529,0.0001947799,0.3441586],"study_design_scores_gemma":[0.000003880748,0.00001251591,0.000003188161,0.0001205686,0.000009220938,0.000001041172,0.00004749011,0.56857,0.2826784,0.1484524,0.00004178498,0.00005963842],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01213383,0.0001240696,0.9839656,0.0006318788,0.00009899805,0.0003436037,0.00000447617,0.0003168447,0.00238068],"genre_scores_gemma":[0.9800313,0.000091147,0.01959129,0.00001105938,0.000060911,0.00006680792,6.386339e-7,0.00002044915,0.0001263814],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9678975,"threshold_uncertainty_score":0.3867982,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06288092941123159,"score_gpt":0.2993721743013264,"score_spread":0.2364912448900948,"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."}}