{"id":"W3175980437","doi":"10.3390/rs13132460","title":"RUF: Effective Sea Ice Floe Segmentation Using End-to-End RES-UNET-CRF with Dual Loss","year":2021,"lang":"en","type":"article","venue":"Remote Sensing","topic":"Arctic and Antarctic ice dynamics","field":"Earth and Planetary Sciences","cited_by":49,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"ArcticNet","keywords":"Segmentation; Sea ice; Computer science; Artificial intelligence; Synthetic aperture radar; Geology; Conditional random field; Sea ice concentration; Arctic ice pack; Remote sensing; Sea ice thickness; Climatology","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":[],"consensus_categories":[],"category_scores_codex":[0.0002160332,0.0001922241,0.0002164538,0.00007335693,0.0003727113,0.0001032923,0.00005330341,0.0000699897,0.000162654],"category_scores_gemma":[0.00006998041,0.0001682818,0.00005264188,0.0004129807,0.00008887504,0.0002116066,0.00002319372,0.0002157023,0.00008084845],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004076344,"about_ca_system_score_gemma":0.0001550535,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004225186,"about_ca_topic_score_gemma":0.002582839,"domain_scores_codex":[0.998432,0.0001907031,0.0001898225,0.0003996661,0.0003803105,0.0004074512],"domain_scores_gemma":[0.9991614,0.0002683026,0.0000857952,0.0002050933,0.0001168175,0.0001625616],"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":[0.0002708385,0.00001528601,0.03028779,0.00009775878,0.0001499951,0.001968218,0.002854848,0.06202434,0.005911181,0.00001312095,0.00003907685,0.8963675],"study_design_scores_gemma":[0.0007276559,0.0002020005,0.04774137,0.0003261896,0.0001514892,0.002124485,0.002706156,0.9405032,0.003855747,0.0002476281,0.0008876775,0.0005263592],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9476717,0.00005871513,0.04907387,0.0003181248,0.0003419001,0.0002124071,0.0000279346,0.0000461065,0.002249212],"genre_scores_gemma":[0.9074914,0.00002136711,0.09135799,0.0005565875,0.000257372,5.334982e-9,0.0001535764,0.00001106489,0.0001506574],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8958412,"threshold_uncertainty_score":0.6862333,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01086140988049485,"score_gpt":0.2294899465140868,"score_spread":0.2186285366335919,"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."}}