| Title |
Gestų atpažinimo sistemos InMoov robotui sukūrimas ir tyrimas / |
| Translation of Title |
Development and research of gesture recognition system for InMoov robot. |
| Authors |
Pilelis, Aivaras |
| Full Text |
|
| Pages |
50 |
| Keywords [eng] |
neural network ; cnn ; kinect ; blazePose ; inmoov |
| Abstract [eng] |
This project development of an algorithm for the inmoov robot to recognize and reproduce gestures using the latest artificial intelligence model to track human posture in a 3d coordinate system is presented. the work introduces the inmoov robot equipment base and capabilities. an overview of recent work in the field of human robot interaction is provided. three methods for obtaining 3d coordinates of human joints are analyzed: blazepose, mobilehumanpose, residualpose. in paper author proposed combined method consisting of a neural network model for determining 2D coordinates and an analytical algorithm for transferring the obtained 2d points to 3d coordinate system. The research method used in the work is literature analysis, generalization and testing of algorithms, presentation of results. |
| Dissertation Institution |
Kauno technologijos universitetas. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2022 |
