Key points are not available for this paper at this time.
Abstract: Sign language is one of the oldest and most natural forms of language communication. These languages are essentially designed to support deaf people. However, since most people do not understand sign language and it is difficult to find an interpreter, we have developed a real-time method for fingerspelling using neural networks in sign language. In our approach, the hand first passes through the filter and, after applying the filter, passes through a classifier that predicts the gesture category. Our method provides higher accuracy for the 26 letters of the alphabe
Singh et al. (Wed,) studied this question.