Liliana, and Chae, Ji-hun and Lee, Joon-Jae and Lee, Byung-Gook (2020) A robust method for VR-based hand gesture recognition using density-based CNN. [UNSPECIFIED]
PDF Download (496Kb) | |
PDF Download (2943Kb) |
Abstract
Many VR-based medical purposes applications have been developed to help patients with mobility decrease caused by accidents, diseases, or other injuries to do physical treatment efficiently. VR-based applications were considered more effective helper for individual physical treatment because of their lowcost equipment and flexibility in time and space, less assistance of a physical therapist. A challenge in developing a VR-based physical treatment was understanding the body part movement accurately and quickly. We proposed a robust pipeline to understanding hand motion accurately. We retrieved our data from movement sensors such as HTC vive and leap motion. Given a sequence position of palm, we represent our data as binary 2D images of gesture shape. Our dataset consisted of 14 kinds of hand gestures recommended by a physiotherapist. Given 33 3D points that were mapped into binary images as input, we trained our proposed density-based CNN. Our CNN model concerned with our input characteristics, having many blank block pixels, single-pixel thickness shape and generated as a binary image. Pyramid kernel size applied on the feature extraction part and classification layer using softmax as loss function, have given 97.7% accuracy.
Item Type: | UNSPECIFIED |
---|---|
Additional Information: | terlambat mengajukan IKP2M karena menunggu proses indexing di scopus |
Uncontrolled Keywords: | 2D image gesture representation, Binary image learning, Density-based CNN, Hand gesture recognition, VR-based physical treatment |
Subjects: | Q Science > QA Mathematics > QA76 Computer software |
Divisions: | Faculty of Industrial Technology > Informatics Engineering Department |
Depositing User: | Admin |
Date Deposited: | 12 May 2020 02:30 |
Last Modified: | 12 Aug 2020 19:49 |
URI: | https://repository.petra.ac.id/id/eprint/20271 |
Actions (login required)
View Item |