Intelligent Sign Language Recognition Enhancing Accessibility for the Deaf Community

Every human being is unique in their own way while some suffer from disabilities like hearing and speech. It is unfair for individuals to encounter hardships when engaging in a fundamental activity such as having a conversation with others. It is commonly understood that having a conversation with people having such kinds of disabilities is a challenging task. Although various sign language standards exist, many individuals are reluctant to invest the additional effort required to learn these to facilitate effective communication with individuals with disabilities. This is where Sign language detection systems come in, they offer a solution to this problem by empowering individuals with disabilities to actively participate in conversations without necessitating knowledge of sign language. In this study, we propose a multilayer Convolutional Neural Network (CNN) model designed to identify ten different signs of American Sign Language (ASL). The model is trained and evaluated on a comprehensive dataset, achieving a high validation accuracy of 95%. The CNN architecture enables the accurate recognition and classification of hand gestures associated with ASL, empowering individuals with disabilities to effectively communicate with others.