Wearable Glove: Sign Language Interpretation with AI-Enabled Finger - Mounted Sensors
International Research Journal of Multidisciplinary Technovation · 18.07.2026 · 250-265
DOI: 10.54392/irjmt26415 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Research Journal of Multidisciplinary Technovation, online erschienen am 18.07.2026. Verfasst von Valarmathi R und 2 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Understanding sign language can facilitate communication between deaf people and hearing people. Nevertheless, a considerable communication gap still exists between sign language users and those who are unfamiliar with its gestures. To address this challenge, the present study combines sensor technology with machine learning (ML) algorithms to develop an AI-enabled wearable glove (AIEWG) for real-time American Sign Language (ASL) recognition and text translation. Two MPU6050 inertial measurement units and three flex sensors were used to capture finger-bending movements, hand motions, and wrist orientations. The proposed system recognizes individual letters (A–Z) and expressions such as "hello," "help," "thank you," "me," and "please." The collected sensor readings were preprocessed for machine learning analysis to interpret static gestures. Several ML and DL models were employed, among which the Particle Swarm Optimization (PSO)-tuned Extra Trees classifier (EXT) demonstrated superior performance, with precision and F1-score values of 99.87%, an AUC-ROC of 99.9%, and recall and accuracy of 99.86%. For real-time deployment, the AIEWG captures sensor data, which are then classified using a PSO-tuned EXT model deployed on an ESP32. The trained model is embedded in the firmware as a C-compatible (.h) file to enable real-time inference in the Arduino IDE. The system operates at a sampling frequency of 100 Hz, with each gesture captured in a continuous 3-second window. The recognized gestures are converted into text and transmitted via Wi-Fi to a web server, where they are displayed through an accessible graphical user interface (GUI). This enables uninterrupted interaction between deaf users and people who are unfamiliar with sign language.
Zitieren
APA
R, V., K, G. B., & R.D, M. (2026). Wearable Glove: Sign Language Interpretation with AI-Enabled Finger - Mounted Sensors. International Research Journal of Multidisciplinary Technovation, 250-265. https://doi.org/10.54392/irjmt26415
Vancouver
R V, K GB, R.D M. Wearable Glove: Sign Language Interpretation with AI-Enabled Finger - Mounted Sensors. International Research Journal of Multidisciplinary Technovation. 2026:250-265. doi:10.54392/irjmt26415
BibTeX
@article{R2026,
author = {R, Valarmathi and K, Ghousiya Begum and R.D, Manoj},
title = {Wearable Glove: Sign Language Interpretation with AI-Enabled Finger - Mounted Sensors},
journal = {International Research Journal of Multidisciplinary Technovation},
year = {2026},
pages = {250--265},
doi = {10.54392/irjmt26415}
}
EUVIMED