Sign Language Recognition using Deep Learning

Sign language recognition is extremely pivotal for filling the void existing between those deaf or hard of hearing, and the rest of mankind. Computer vision advancement through deep learning techniques has been followed successfully with remarkable success across images. This research paper presents a comprehensive study on the application of CNN for sign language recognition. The primary objective is to develop an accurate and efficient system to recognize sign language gestures and translate them into corresponding text or speech. The proposed approach comprises a deep learning model trained as 3000 images collected from various datasets and online resources. The proposed deep learning model is trained in such way that it achieves the seem less efficiency on real time images that contains sign language on different scales and poses. The various outperformed metrics such as precision, recall and f1-score are also recorded. The model achieves 0.995%, 0.994%, and 0.992% accuracy during the testing session.