{"id":"W4415845572","doi":"10.1016/j.patcog.2025.112687","title":"Enhancing spatio-temporal zero-shot action recognition with language-driven description attributes","year":2025,"lang":"en","type":"article","venue":"Pattern Recognition","topic":"Human Pose and Action Recognition","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Institute for Information and Communications Technology Promotion; Ministry of Science and ICT, South Korea; Information Technology Research Centre; Korea University","keywords":"Ambiguity; Class (philosophy); Focus (optics); Salient; Adaptability; Process (computing); Action (physics)","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"],"consensus_categories":[],"category_scores_codex":[0.000313907,0.0003020382,0.0002620802,0.0005505491,0.0003662568,0.0004781279,0.0002595572,0.0001696423,0.0002789071],"category_scores_gemma":[0.00005349104,0.000299233,0.0001113281,0.0006770687,0.00004270667,0.002186384,0.00007500427,0.0003234795,0.0007505128],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002050297,"about_ca_system_score_gemma":0.00009324618,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003279002,"about_ca_topic_score_gemma":0.001455687,"domain_scores_codex":[0.9978878,0.0002109507,0.0004771756,0.0006677018,0.0003584485,0.0003978861],"domain_scores_gemma":[0.9987494,0.0001088252,0.0002993527,0.0003501965,0.0003915367,0.0001006682],"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":[0.00008816065,0.0002428237,0.004103501,0.0001873134,0.00009839151,0.00003706762,0.0005652673,0.00001439208,0.01513586,0.00003999004,0.001298444,0.9781888],"study_design_scores_gemma":[0.006210865,0.001217796,0.03241864,0.003806345,0.0004602606,0.0003266246,0.001783543,0.02451934,0.9003482,0.02096898,0.005337992,0.002601335],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4427952,0.00002597958,0.5543417,0.0003931548,0.0005306214,0.0003409992,0.00003520944,0.0004030705,0.001134065],"genre_scores_gemma":[0.9883851,0.00005730366,0.00788086,0.001054729,0.000222935,0.0001851667,0.001858696,0.00002484327,0.0003303542],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9755874,"threshold_uncertainty_score":0.999946,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0576511750919326,"score_gpt":0.2786318774689936,"score_spread":0.220980702377061,"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."}}