{"id":3242,"date":"2022-07-13T09:30:00","date_gmt":"2022-07-13T07:30:00","guid":{"rendered":"https:\/\/cirpicme.org\/?page_id=3242"},"modified":"2022-07-26T13:05:44","modified_gmt":"2022-07-26T11:05:44","slug":"potentials-of-few-shot-learning-for-quality-monitoring-in-laser-welding-of-hairpin-windings","status":"publish","type":"page","link":"https:\/\/cirpicme.org\/index.php\/forming-welding-2\/potentials-of-few-shot-learning-for-quality-monitoring-in-laser-welding-of-hairpin-windings\/","title":{"rendered":"Potentials of Few-Shot Learning for Quality Monitoring in Laser Welding of Hairpin Windings"},"content":{"rendered":"\n

by Tim Raffin, Andreas Mayr, Marcel Baader, Nadine Laube, Alexander Kuehl, Joerg Franke (Germany)<\/em><\/p>\n\n\n\n

<\/p>\n\n\n\n

Abstract<\/strong><\/p>\n\n\n\n

For analysis of high-dimensional data in electromechanical manufacturing, data-driven techniques such as Deep Learning show great potential. However, the vast amount of required data poses a major barrier to the industrial adoption of said techniques. In hairpin stator production especially, due to its novelty, short innovation cycles further amplify this situation. In contrast to classical Deep Learning approaches, Few-Shot Learning aims at generalizing well despite only limited available data. Therefore, this paper presents an empirical evaluation of Few-Shot Learning techniques in the context of hairpin stator production. Accordingly, we evaluate and benchmark several Few-Shot Learning architectures on a dataset for laser welding of hairpins.<\/p>\n\n\n\n

Keywords<\/strong>: Few-Shot Learning; Deep Learning; Hairpin technology; Electric drives production; Quality monitoring<\/p>\n\n\n\n

<\/div>\n\n\n\n

Video presentation<\/strong><\/p>\n\n\n\n