Maldivian Sign Language Recognition Application
Randha Adam, Adam Khalid · 2022 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) · 2022
Maldivian Sign Language (MVSL) is the national sign language for the deaf or the hearing-impaired people in Maldives. These people with disabilities could easy communicate with other deaf people but face great difficulty conversing with ordinary people. There are many recognition applications and models made for many sign languages with different approaches. Sadly, no such technology has been made to recognize MVSL signs. Therefore, this research aims to recognize three dynamic MVSL signs (“water”, “promise” and “read”) in real-time and later expand to recognize more signs to create a translation application for MVSL. We propose an approach with MediaPipe holistic to extract key point landmarks from the face, two hands and the pose from the videos and build the model for recognition using LSTM. The model was able to obtain an accuracy of 93%.