{"id":"W3158216587","doi":"10.3390/s21092984","title":"Quantization and Deployment of Deep Neural Networks on Microcontrollers","year":2021,"lang":"en","type":"article","venue":"Sensors","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":188,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute","funders":"Centre National de la Recherche Scientifique; Agence Nationale de la Recherche","keywords":"Computer science; Microcontroller; Quantization (signal processing); Artificial neural network; Deep learning; Software deployment; MNIST database; Inference; Embedded system; Artificial intelligence; Computer engineering; Computer hardware; Algorithm","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004045385,0.0008408701,0.0003328669,0.0005139849,0.000188674,0.0006309376,0.001593886,0.0003248682,0.007972012],"category_scores_gemma":[0.002288085,0.0002865018,0.0002449186,0.0003641224,0.0003336259,0.001337381,0.0008269204,0.0008270303,0.001144168],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008465511,"about_ca_system_score_gemma":0.000615429,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002593079,"about_ca_topic_score_gemma":0.002970281,"domain_scores_codex":[0.9994841,0.00007534376,0.00004649353,0.0001182096,0.0002100923,0.00006580367],"domain_scores_gemma":[0.9993954,0.0001808191,0.00005364836,0.0001287743,0.0002082763,0.00003320801],"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":[0.0007499409,0.0001800684,0.002427886,0.0006903511,0.00009962604,0.0005825898,0.0002943719,0.2566348,0.09373979,0.01748024,0.0250206,0.6020997],"study_design_scores_gemma":[0.00009072095,0.000476438,0.001706349,0.00008199621,0.0000421787,0.0002256145,0.0001068441,0.8336893,0.1335966,0.008501443,0.02143701,0.00004537135],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1091315,0.001056975,0.848026,0.0006204861,0.0003397277,0.0003079331,0.0008198138,0.02743908,0.01225832],"genre_scores_gemma":[0.7738956,0.0003991299,0.2173453,0.0002713526,0.00003363699,0.0002582931,0.0008539952,0.0005720364,0.006370615],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007972012,"threshold_uncertainty_score":0.02666903,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01137228475476032,"score_gpt":0.242676252022297,"score_spread":0.2313039672675367,"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."}}