{"id":"W4408345963","doi":"10.1109/icassp49660.2025.10889667","title":"Face-StyleSpeech: Enhancing Zero-shot Speech Synthesis from Face Images with Improved Face-to-Speech Mapping","year":2025,"lang":"en","type":"article","venue":"","topic":"Face recognition and analysis","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Face (sociological concept); Computer science; Speech recognition; Speech synthesis; Zero (linguistics); Shot (pellet); Artificial intelligence; Computer vision; Linguistics","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"],"consensus_categories":[],"category_scores_codex":[0.000455833,0.0004946505,0.0006651323,0.0006707854,0.0003316967,0.0007942041,0.001440729,0.000167972,0.0003793126],"category_scores_gemma":[0.0002292945,0.0004101271,0.0002380508,0.002127079,0.00006656677,0.0006684714,0.0005996106,0.0003304277,0.0006917549],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001528175,"about_ca_system_score_gemma":0.0001834969,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007831333,"about_ca_topic_score_gemma":0.0004360903,"domain_scores_codex":[0.9966938,0.0001432384,0.0005844019,0.001295423,0.0005417259,0.0007413432],"domain_scores_gemma":[0.9976646,0.0005259928,0.0001638776,0.00114718,0.0002072475,0.0002911157],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00004446448,0.0002797122,0.0005341325,0.0001040496,0.0008151794,0.000110709,0.001848448,0.001004446,0.5020874,0.001057882,0.00457728,0.4875364],"study_design_scores_gemma":[0.0004708365,0.00005020363,0.0004933533,0.0003322366,0.00009405329,0.000009209225,0.002357171,0.05309085,0.9386396,0.000629447,0.003086125,0.0007468935],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04141876,0.000104793,0.9383764,0.006734262,0.0001794137,0.0004560734,0.00003969165,0.0006571548,0.01203343],"genre_scores_gemma":[0.7129059,0.00005032833,0.2712492,0.001982607,0.00004080807,0.00007394139,0.00001302213,0.00002964748,0.01365454],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6714871,"threshold_uncertainty_score":0.9998351,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01777450166422592,"score_gpt":0.2459365036945091,"score_spread":0.2281620020302831,"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."}}