HandSpeak: A Sign Language Learning Platform (ISL)

Indian Sign Language (ISL) is crucial in providing an effective communication mode for those who have hearing and speech difficulties. In most cases, it becomes hard to learn ISL due to non-availability of interactive means of practice. As a result, we have designed HandSpeak: A Sign Language Learning Platform, which involves the implementation of a vision-based sign language module along with a gesture recognition tool to help users learn effectively. This system makes use of techniques such as computer vision, MediaPipe for hand landmarks, and a Convolutional Neural Network for sign gesture classification. Users can practice different sign gestures using their cameras and receive real-time feedback in the form of text and speech output. Moreover, it supports quiz-based learning as a way of improving the process. In the experimentation stage, our MediaPipe-CNN method performs well compared with existing methods such as video-based learning and CNN-based learning.