{"id":"W4403102908","doi":"10.1007/s10032-024-00499-2","title":"Unpaired document image denoising for OCR using BiLSTM enhanced CycleGAN","year":2024,"lang":"en","type":"article","venue":"International Journal on Document Analysis and Recognition (IJDAR)","topic":"Image and Signal Denoising Methods","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Artificial intelligence; Computer science; Pattern recognition (psychology); Image (mathematics); Noise reduction; 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":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.00145718,0.0002469519,0.0003237545,0.001372803,0.0003184533,0.003848292,0.0005163404,0.00007390088,0.0002111678],"category_scores_gemma":[0.00009003589,0.0002011615,0.0005356576,0.0006579788,0.00004890889,0.001635473,0.0001151526,0.0002671535,0.00002886613],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000274052,"about_ca_system_score_gemma":0.00009733673,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003179956,"about_ca_topic_score_gemma":0.000005891835,"domain_scores_codex":[0.9975051,0.0002082291,0.0006575077,0.0005271803,0.0007950551,0.0003069259],"domain_scores_gemma":[0.9985569,0.0003174462,0.0002535239,0.0001808645,0.000517746,0.0001735173],"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":[0.0003390327,0.0002149016,0.00008660534,0.00005861455,0.006784728,0.0005482606,0.001372249,0.001353529,0.08863408,0.004635962,0.0009802766,0.8949918],"study_design_scores_gemma":[0.005260757,0.0008543258,0.0008163401,0.001663811,0.003380727,0.0009170812,0.0003408089,0.3335301,0.3325384,0.3057003,0.01313383,0.001863506],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1231826,0.0003932663,0.8723394,0.001892325,0.001598636,0.0001431,0.00001232111,0.00007687199,0.0003615379],"genre_scores_gemma":[0.8498691,0.0004599954,0.1470335,0.0008291395,0.0009414623,0.00001611321,0.00004606017,0.00002606133,0.0007786396],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8931283,"threshold_uncertainty_score":0.9971858,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02578919183659872,"score_gpt":0.3390521256032745,"score_spread":0.3132629337666758,"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."}}