Back close

Sign Language Classification With MediaPipe Hand Landmarks

Publication Type : Conference Proceedings

Publisher : IEEE

Source : 2023 International Conference on Energy, Materials and Communication Engineering (ICEMCE)

Url : https://doi.org/10.1109/icemce57940.2023.10434034

Campus : Nagercoil

School : School of Computing

Year : 2023

Abstract : The research presented here offers a strategy for addressing the communication issues that the deaf-mute community face on a regular basis. The suggested method uses the open-source MediaPipe framework and machine learning algorithms in an effort to simplify Sign Language Recognition (SLR) by utilizing recent developments in artificial intelligence. The created prediction model is intended to be portable and adaptable to smart devices, allowing for easy integration into regular communication tools. Two datasets—the American Sign Language number system and American Sign Language gestures were used to assess the framework's capabilities. The model's astounding average accuracy of 97% demonstrates the Support Vector Machine (SVM) algorithm's dominance over other machine learning techniques in terms of effectiveness and accuracy.

Cite this Research Publication : R. Saravanan, S. Veluchamy, Sign Language Classification With MediaPipe Hand Landmarks, 2023 International Conference on Energy, Materials and Communication Engineering (ICEMCE), IEEE, 2023, https://doi.org/10.1109/icemce57940.2023.10434034

Admissions Apply Now