{"id":"W2078914384","doi":"10.1109/icdar.2011.43","title":"Novel Data Representation for Text Extraction from Multispectral Historical Document Images","year":2011,"lang":"en","type":"article","venue":"","topic":"Image Retrieval and Classification Techniques","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; McGill University","keywords":"Multispectral image; Computer science; Artificial intelligence; Pattern recognition (psychology); Feature extraction; Classifier (UML); Pixel; Grayscale; Support vector machine; Visualization; Computer vision","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0004138144,0.0005583032,0.0007687184,0.001904026,0.0002933969,0.0009607124,0.0008748348,0.000726357,0.002272464],"category_scores_gemma":[0.001272136,0.0001718771,0.0006326459,0.001975497,0.0002436518,0.001687212,0.0005057515,0.0005666877,0.002268228],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003332057,"about_ca_system_score_gemma":0.0003969349,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001053699,"about_ca_topic_score_gemma":0.001218344,"domain_scores_codex":[0.9996142,0.00004132482,0.0000622872,0.0000918043,0.000150904,0.00003944983],"domain_scores_gemma":[0.9995356,0.0000971087,0.00006738934,0.0001027949,0.0001751083,0.00002192179],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002678081,0.0001487018,0.000792735,0.0003085919,0.00003569647,0.0001900679,0.00009013537,0.004639958,0.163424,0.003052239,0.004547391,0.8225029],"study_design_scores_gemma":[0.00009145388,0.0005427423,0.009664763,0.000114373,0.000176992,0.002240945,0.0002724428,0.5998734,0.3275629,0.006131825,0.05319579,0.0001323106],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02462851,0.0007785815,0.96887,0.0001765063,0.0001246791,0.0001349066,0.001378055,0.002986407,0.0009224526],"genre_scores_gemma":[0.1440353,0.0008771436,0.8453214,0.00009102037,0.0001210578,0.0004076619,0.005477546,0.000144494,0.003524326],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002272464,"threshold_uncertainty_score":0.007602155,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1589893034306319,"score_gpt":0.3497189274167667,"score_spread":0.1907296239861348,"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."}}