{"id":"W3136599387","doi":"10.1109/access.2021.3096823","title":"uTHCD: A New Benchmarking for Tamil Handwritten OCR","year":2021,"lang":"en","type":"preprint","venue":"IEEE Access","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Trent University; Nottingham Trent University","keywords":"Computer science; Tamil; Database; Grid; Benchmark (surveying); Benchmarking; Font; Optical character recognition; Artificial intelligence; Scripting language; Metadata; Field (mathematics); Information retrieval; Natural language processing; Pattern recognition (psychology); World Wide Web; Image (mathematics)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0005211202,0.0005120614,0.0006945136,0.000382943,0.0002034274,0.003534223,0.004684251,0.0005532216,0.0001132922],"category_scores_gemma":[0.0001203351,0.0005484009,0.000451428,0.000473711,0.0000496453,0.001240308,0.002962906,0.0007337037,0.00001930772],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001479407,"about_ca_system_score_gemma":0.001011221,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005964561,"about_ca_topic_score_gemma":0.0001708104,"domain_scores_codex":[0.9965621,0.0001163411,0.0006559825,0.001500093,0.0005339718,0.0006315309],"domain_scores_gemma":[0.9968482,0.0002635595,0.0004413829,0.00165011,0.0005193019,0.0002775023],"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":[0.00001923144,0.0001869349,0.0003710349,0.0007884157,0.0002457812,0.0001205987,0.001093904,0.0001385697,0.001452253,0.001284671,0.1183897,0.8759089],"study_design_scores_gemma":[0.003421952,0.0004521169,0.00237588,0.006034564,0.0003799799,0.0002186873,0.000101403,0.08483793,0.5197091,0.2937652,0.0825808,0.006122495],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005868091,0.0005893893,0.9848798,0.001239443,0.003456656,0.001138941,0.00003869514,0.001005861,0.001783135],"genre_scores_gemma":[0.3387715,0.0006455379,0.6488537,0.003972436,0.003783837,0.00155889,0.0003178797,0.0001474292,0.001948818],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8697864,"threshold_uncertainty_score":0.9996967,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06194948546271989,"score_gpt":0.3390967646218885,"score_spread":0.2771472791591686,"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."}}