{"id":"W4407376597","doi":"10.1109/tpami.2025.3541028","title":"Pixel-Inconsistency Modeling for Image Manipulation Localization","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Pattern Analysis and Machine Intelligence","topic":"Image Processing Techniques and Applications","field":"Engineering","cited_by":47,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"Nanyang Technological University; Agency for Science, Technology and Research","keywords":"Artificial intelligence; Computer vision; Pixel; Computer science; Image manipulation; Image (mathematics); Image processing; Pattern recognition (psychology)","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":[0.0000900639,0.0001331722,0.0001635577,0.0003618758,0.0001900879,0.00007348403,0.00008820865,0.00005534336,0.00002955272],"category_scores_gemma":[0.00000230816,0.0001337531,0.0001235209,0.0005929683,0.00002258933,0.0001037089,0.000001098395,0.0001002812,0.000002930979],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000338785,"about_ca_system_score_gemma":0.00000706112,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001112645,"about_ca_topic_score_gemma":0.0002451283,"domain_scores_codex":[0.9993153,0.000009318203,0.0002756879,0.0002136533,0.00006575031,0.0001203337],"domain_scores_gemma":[0.9996397,0.00003555157,0.0000252616,0.0001849892,0.00008361157,0.00003082869],"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":[0.000003778973,0.00003250665,0.00003202077,0.00007834532,0.0001521702,1.572982e-7,0.00004175546,0.639704,0.0009654484,0.0001195137,0.00001268801,0.3588576],"study_design_scores_gemma":[0.00003813789,0.00001056816,0.00001029385,0.00002873651,0.0004089137,6.881609e-7,0.00002144995,0.9327534,0.06486423,0.001665301,0.00008197854,0.0001163274],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0004932488,0.0001182451,0.9985229,0.0001222019,0.00006015015,0.0001880008,0.00003249761,0.0002394621,0.0002233295],"genre_scores_gemma":[0.9905172,0.0002720426,0.008802551,0.0001213681,0.000008854735,0.0001233562,0.00002321053,0.00001439331,0.0001170152],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.990024,"threshold_uncertainty_score":0.5454293,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01968662579945709,"score_gpt":0.2821643964836529,"score_spread":0.2624777706841959,"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."}}