Title: A wearable glove model with custom PCB implementation for Malaysian Sign Language recognition

Authors: Zinah Raad Saeed; Zurinahni Binti Zainol; Ibrahim Ismael Alnaib; Karam Khairullah Mohammed

Addresses: School of Computer Sciences, Universiti Sains Malaysia, Penang, Malaysia ' School of Computing and Informatics, Albukhary International University, Kedah, Malaysia ' Department of Electrical Engineering, College of Engineering, University of Mosul, 41001, Mosul, Iraq ' Department of Electrical Engineering, College of Engineering, University of Mosul, 41001, Mosul, Iraq

Abstract: An important means of communication for those with hearing loss is sign language. However, the limited prevalence of sign language knowledge in society leads to a communication gap within the deaf and hard-of-hearing community. Developing glove-based interaction devices was prohibitively expensive and fraught with technical issues, notably the need for users to wear cumbersome gloves with numerous internal components, resulting in discomfort and restricted hand movement. To capture finger and hand movements, this paper initially designed an inexpensive and effective data glove with fewer wiring connections and a custom printed circuit board (PCB). Secondly, a novel dataset was collected comprising 20 dynamic Malaysian Sign Language recognition (MSL) words used in daily life. The hybrid CNN-LSTM algorithm was implemented, CNN was utilised for deep feature extraction, and LSTM was used for recognising Malaysian dynamic sign language. In contrast, the results illustrated the effectiveness of the CNN-LSTM method in recognising Malaysian dynamic sign language, paving the way for advancements in gesture recognition technology and benefiting a wide range of human-computer interaction and assistive applications.

Keywords: printed circuit board; PCB; glove; sensors; sign language; conventional neural network; CNN; long short-term memory network; LSTM.

DOI: 10.1504/IJPEC.2026.155136

International Journal of Power and Energy Conversion, 2026 Vol.17 No.3, pp.306 - 320

Received: 23 Apr 2024
Accepted: 02 Jun 2024
Published online: 28 Jul 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article