Sign Language Detection Using Deep Learning

In our digital age, communication should be accessible for everyone. Communication is a fundamental human right, yet millions of individuals who use sign language face daily challenges in expressing themselves and understanding others. Sign Language Detection and Text Conversion will help to break down the walls that make communication difficult for sign language users. This paper presents an innovative solution for this problem where we used MobileNetV2, Resnet50 and EfficientNet-B0 algorithm on a Kaggle dataset containing American Sign Language (ASL). We were able to achieve 98.96% accuracy using EfficientNet-B0, 98.33% accuracy with RestNet50 and 98.12% for MobileNetV2. Models were trained using 55,680 images, 13,920 images were used for validation and 17,400 images for testing purpose. TensorFlow framework is implemented for MobileNetV2, ResNet50, and EfficientNet-B0. The network consists of several layers like global average pooling, dense layer with 128 units, ReLU activation function with 29 units and a softmax with 5 epochs. Importantly, we convert these gestures into alphabets, making communication more reachable.