Worked on hand gestures and recognizing the words and combine them to form a meaningful sentences. I have used two machine learning models - SVM and LSTM, to predict the sign. The models were trained and tested using 42 different sign words and 28 sentences. LSTM outperformed with accuracy 98.21 %.
venkateshbh99 / continuous-sign-language-using-leap-motion-sensor Goto Github PK
View Code? Open in Web Editor NEWWorked on hand gestures and recognizing the words and combine them to form a meaningful sentences. I have used two machine learning models - SVM and LSTM, to predict the sign. The models were trained and tested using 42 different sign words and 28 sentences. LSTM outperformed with accuracy 98.21 %.