Sign to Speak: Real-time Recognition for Enhance Communication

Recent developments in the area of research on sign language recognition have led to making use of deep learning techniques with the goal to assist the self-learning transition of sign language from one language to another. In order to detect and classify characters from photographs in multiple sign languages, a number of deep-learning models, have been deployed. SLR is a prominent area of study and technology that aims to mitigate speaking barriers for individuals who unable to talk and listen. This involves the development of tools capable of understanding and deciphering sign language movements, hence facilitating a wider range of interactions for those who are difficult of listening and speaking. This paper demonstrates a novel model called CNN Inception Net, which is a neural network based on convolution. The use of an Indian sign dataset enables the achievement of a notable performance rate of 98.9%, enabling precise recognition of actions in the language of signs.