An Integrated Transfer Learning Model for American Sign Language Recognition

American Sign Language (ASL) is a visual language used by individuals with speech and hearing disabilities to communicate effectively in their daily interactions. Recognizing the importance of improving accessibility and inclusivity, efforts have been directed towards developing a robust model to recognize ASL signs using advanced deep learning techniques. In existing research, fixed gestures were used to train the model. In this research, the dynamic gestures were used to identify the ASL. The ASL recognition system employs Data Augmentation techniques to enhance the diversity and quality of the training dataset. At the core of the system is a Convolutional Neural Network (CNN) architecture, specifically the Inception V3 model, integrated with Transfer Learning. The Inception V3 model is trained on a diverse dataset of ASL signs, it becomes capable of interpreting gestures from both images and video feeds with high accuracy and efficiency.