AMERICAN SIGN LANGUAGE DETECTION WITH CONVOLUTIONAL NEURAL NETWORK

Abstract

Humans are social creatures, one of the ways humans socialize is by communicating with each other. Socializing is needed for humans to be able to adapt, recognize and be able to move towards their environment. However, this is an obstacle for the deaf community, even though the presence of sign language can help to socialize, but not everyone can understand it. This study will propose the detection of sign language so that it is easy to understand, the sign language that will be detected is American Sign Language (ASL). American Sign Language is used because it is quite well known for its use in terms of sign language research and will use the Machine Learning method, namely Convolutional Neural Network (CNN). The use of this method itself was chosen because it has a high level of accuracy, in this study the results of the resulting accuracy rate were 99.89%.

Keyword: Communication, ASL, CNN, Detection, Accuracy.