{"id":"W4319300892","doi":"10.1109/wacv56688.2023.00278","title":"TeST: Test-time Self-Training under Distribution Shift","year":2023,"lang":"en","type":"article","venue":"2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":26,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Test data; Artificial intelligence; Test (biology); Machine learning; Inference; Adaptation (eye); Data mining; 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":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0006684003,0.0003720844,0.0004343023,0.0004177486,0.0003316267,0.0004358462,0.001667316,0.0001689076,0.0001585709],"category_scores_gemma":[0.00006545962,0.000372931,0.0002009288,0.001680489,0.0001340213,0.0005058583,0.0004356172,0.0003979605,0.003454026],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000940719,"about_ca_system_score_gemma":0.0002044483,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000009063944,"about_ca_topic_score_gemma":0.000003614661,"domain_scores_codex":[0.9968937,0.0001274371,0.0007653225,0.0009515458,0.0007105226,0.0005514936],"domain_scores_gemma":[0.996604,0.001155758,0.0003888927,0.001225501,0.0003718981,0.0002539299],"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":[0.00006327468,0.001882556,0.0006007009,0.0001727102,0.000220147,0.00003738708,0.006060983,0.007646605,0.01860883,0.3615602,0.07742597,0.5257206],"study_design_scores_gemma":[0.001005191,0.0007370574,0.01926312,0.0002707957,0.00002610715,0.00001834996,0.0001845312,0.9275962,0.001590718,0.01253326,0.03604495,0.0007296833],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003700806,0.000008455458,0.9856916,0.005626357,0.0003261119,0.0006363468,0.00009523856,0.001177543,0.002737527],"genre_scores_gemma":[0.9686872,0.00002902842,0.02833139,0.0007391247,0.0002218641,0.0001994704,0.0002278939,0.00003790836,0.001526121],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9649864,"threshold_uncertainty_score":0.9998723,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02988191807964998,"score_gpt":0.2922269165930714,"score_spread":0.2623449985134214,"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."}}