Finger Gesture Recognition using Recurrent Neural Network and Convolution Neural Network

Communication is essential for people to interact with one another. Sadly, not everyone can communicate verbally, so they resort to Sign language to engage in conversations and express themselves using various hand gestures. The deaf community mostly uses American sign language (ASL). However, only a few people can interpret ASL, and various translation systems have been developed. We aim to implement various Deep Learning techniques to translate ASL and read the sign out loud to find and develop an efficient and accurate sign language translation to make communication more accessible for hard-of-hearing individuals by building Recurrent Neural Networks and Convolution Neural Networks. On evaluating both these models highlighting the strengths and weaknesses of each approach. This is conducted based on various hyperparameters such as the Number of Epochs, Accuracy, Time, and Space Complexity. The study's findings offer exciting prospects for improving hand gesture recognition technology.