I copy human brain (not as good). I look at the person signing using video cameras, then I focus on hands, face, body movements (because they make up sign languages). While someone is signing in the video, I detect individual signs. I identify individual signs by giving them labels (i.e. sign for ‘thank you’). Then a machine knows that the sign in the video is ‘thank you’ and once it sees this sign, the machine should say ‘thank you too’ to be polite to the signer. We use vision processing and machine learning techniques to recognise and understand sign languages
Comments