Automated sign language translation using deep learning

Wong, Jia Kang (2025) Automated sign language translation using deep learning. Final Year Project, UTAR.

PDF Download (6Mb)

Abstract

This project focuses on developing a system for automated static gesture sign language translation using deep learning. With the increasing demand for accessible communication tools, particularly for the hearing-impaired community, the need for reliable sign language translation systems is growing. The main challenge addressed in this project is the recognition and translation of static sign language gestures into text, which is less complex than dynamic gestures involving movement. The methodology involves processing images of static sign language gestures using hand landmark detection with MediaPipe. These landmarks are then normalized and input into a deep learning model, trained on processed dataset images, to predict the corresponding sign. The model architecture consists of multiple dense layers with batch normalization and dropout to ensure robust learning. The system is integrated into a user-friendly application that offers real-time sign language translation through a webcam feed, with features such as dynamic confidence threshold adjustment, translation history tracking, and a sign language dictionary. The results show that the system is capable of accurately recognizing and translating static sign language gestures with high confidence, as validated by the test dataset. The system is efficient, easy to use, and highly adaptable for future enhancements. This project demonstrates the potential of deep learning in bridging communication gaps for the hearing-impaired community and sets the groundwork for future work in dynamic sign language translation.

Actions (login required)