Development of Malaysian sign language (MSL) translator using deep learning approach

Zakaria, Muhammad Mu’izz (2022) Development of Malaysian sign language (MSL) translator using deep learning approach. Project Report. Melaka, Malaysia, Universiti Teknikal Malaysia Melaka. (Submitted)

	Text (24 Pages) Development of Malaysian sign language (MSL) translator using deep learning approach.pdf - Submitted Version Download (851kB)
	Text (Full Text) Development of Malaysian sign language (MSL) translator using deep learning approach.pdf - Submitted Version Restricted to Registered users only Download (1MB)

Abstract

Malaysian Sign Language (MSL) has been used as a means of communication in Malaysia between the deaf and the hearing people. Many people still could not take a grasp on what is being conveyed by the deaf which may cause misunderstanding. Thus, the aim of this project is to develop a MSL translator interface which able to detect the type of gestures when signing and a speech to text conversion. A deep learning approach is used focusing on the image classification that are known as “Hai”, “Tak”, “Terima Kasih” by utilizing TensorFlow and Mediapipe software where the model build will be trained using labeled images and identify its classes. This project has been develop succesfully of a MSL translator interface that can assist both the deaf and hearing people for a better two way communication in the future with the rate of confidence of more than 95 percent. In essence, it is to help in bridging the gap between the communication of the deaf and hearing people.

Actions (login required)

View Item