Szczegóły publikacji
Opis bibliograficzny
Dynamic hand gesture recognition for human-robot collaborative assembly / Bogdan KWOLEK, Sako Shinji // W: Artificial Intelligence and Soft Computing : 22nd International Conference, ICAISC 2023 : Zakopane, Poland, June 18–22, 2023 : proceedings, Pt. 1 / eds. Leszek Rutkowski, Rafał Scherer, Marcin Korytkowski, Witold Pedrycz, Ryszard Tadeusiewicz, Jacek M. Zurada. — Cham : Springer Nature Switzerland, cop. 2023. — (Lecture Notes in Computer Science ; ISSN 0302-9743 ; LNCS 14125. Lecture Notes in Artificial Intelligence). — ISBN: 978-3-031-42504-2; e-ISBN: 978-3-031-42505-9. — S. 112–121. — Bibliogr., Abstr. — Publikacja dostępna online od: 2023-09-14
Autorzy (2)
- AGHKwolek Bogdan
- Shinji Sako
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 151039 |
|---|---|
| Data dodania do BaDAP | 2024-01-22 |
| DOI | 10.1007/978-3-031-42505-9_10 |
| Rok publikacji | 2023 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Wydawca | Springer |
| Konferencja | International Conference on Artificial Intelligence and Soft Computing 2023 |
| Czasopismo/seria | Lecture Notes in Computer Science |
Abstract
In this work, we propose a novel framework for gesture recognition for human-robot collaborative assembly. It permits recognition of dynamic hand gestures and their duration to automate planning the assembly or common human-robot workspaces according to Methods-Time-Measurement recommendations. In the proposed approach the common workspace of a worker and Franka-Emika robot is observed by an overhead RGB camera. A spatio-temporal graph convolutional neural network operating on 3D hand joints extracted by MediaPipe is used to recognize hand motions in manual assembly tasks. It predicts five motion sequences: grasp, move, position, release, and reach. We present experimental results of gesture recognition achieved by a spatio-temporal graph convolutional neural network on real RGB image sequences.
