A New Dataset and Proposed Convolutional Neural Network Architecture for Classification of American Sign Language Digits
This work addresses communication challenges for speech-impaired people, but it is incremental as it builds on existing CNN methods with a new dataset.
The authors tackled the problem of communication barriers for speech-impaired individuals by creating a new American Sign Language digits dataset and proposing a Convolutional Neural Network architecture that achieved 98% test accuracy on this dataset.
According to interviews with people who work with speech impaired persons, speech impaired people have difficulties in communicating with other people around them who do not know the sign language, and this situation may cause them to isolate themselves from society and lose their sense of independence. With this paper, to increase the quality of life of individuals with facilitating communication between individuals who use sign language and who do not know this language, a new American Sign Language (ASL) digits dataset that can help to create machine learning algorithms which need to large and varied data to be successful created and published as Sign Language Digits Dataset on Kaggle Datasets web page, a proposal Convolutional Neural Network (CNN) architecture that can get 98% test accuracy on our dataset presented, and compared with the existing popular CNN models.